Content uploaded by Phil Harris
Author content
All content in this area was uploaded by Phil Harris on Dec 30, 2013
Content may be subject to copyright.
Who sets the
agenda?
1117
Europ ean Jou rnal of Market ing,
Vol. 35 No. 9/10, 2001, pp. 1117-1135.
# MCB Univ ersity Press, 0309-0566
Received June 2000
Revised March 2001
Who sets the agenda?
An analysis of agenda setting and press
coverage in the 1999 Greek Eur opean
elections
Phil Harris and Ioannis Kolovos
The Manchester Metropol itan University , Manchester, UK, and
Andrew Lock
Leeds University Business School, Leeds, UK
Keywords
Marketing, Greece, Politics, Content analysis, Political parties, Media
Abstract
News media have been shown to have a significant influence in the selection and
perception of issues in political campaigns. This has become known as ``agenda-setting’’. The
evolution of the agenda-setting literature is traced and the links with political campaigning and
political marketing are identified. Alth ough the term is widely used in Australasia, Europe and
North America, there is no previous empirical research on agenda-setting in Greece. The article
outlines a content analysis of press coverage over the period of the campaign for the European
elections in Greece in 1999. The results are contrasted with an analysis of party manifestos and
press releases and with public opinion prior to the campaign. Differences between the three
agenda groups are identified.
Introd uction
In this article we present an overview of the literature on agenda-setting and
report on a study made in the context of the 1999 European elections in Greece.
There has been little if any empirical research on agenda -setting in Greece to
date. As Greece has a rather different modern political history and cultur al
tradition s from the USA and the UK, which provid e the bulk of the studies in
the field to date, the study provides a useful compar ative perspective on the
field.
A content ana lysis of six newspapers over the full campa ign period was
carried out. A similar analysis of t he manif estos of six parties was done using
the same categories. In addition, the press releases of the two major parties,
PASOK and New Democracy, were also analysed and categor ised. T hese
results were compared with the public perceptions of the issue pr iorities
identified in a public opinion survey immediat ely prior to the election
campaig n. We were thus able to compare the agendas of the press, the different
parties and the public. Whilst European elections do not necessarily indicate
how people would vote in a general election, they are considered by political
parties in all member states to be a barometer of p ublic opinion and are a valid
context for a content analysis study of a political campaign and press coverage.
The differences reported between the apparent agendas of the press, the
political parties and the public are consistent with the results reported by
T h e r e s e a r c h r e g i s t e r f o r t h i s j o u r n a l i s a v a i l a b l e a t
h t t p : / / w w w . m c b u p . c o m / r e s e a r c h _ r e g i s t e r s
T h e c u r r e n t i s s u e a n d f u l l t e x t a r c h i v e o f t h i s j o u r n a l i s a v a i l a b l e a t
h t t p : / / w w w . e m e r a l d - l i b r a r y . c o m / f t
This article has been independently reviewed by the Editors of EJM.
European
Journal of
Marketing
35,9/10
1118
Harr is e t al. (1999) in respect of the 1997 UK General Election. The implication s
of this for political marketing and political campaign managem ent are
discussed.
Agenda -setting
Lippma nn (1922) argu ed that the mass media are the link between world events
and the pictures of these events in our minds. Without using the modern term,
he is clearly referring to what we have come to call public agenda-setting. In a
similar vein, Cohen (1963) observed that the p ress ``may not be successful much
of the time in telling people what to think, but it is stunningly successful in
telling its r eaders what to think ab out’’.
News media appear to determine to a considerable degree the significant
issues during a political campaign (McCombs and Shaw, 1977). Media coverage
structu res voters’ perceptions of political issues and thus affects the ground on
which campaigns are conducted. This process is commonly described as
``agenda-setting ’’. The term was first used formally in a study of the 1968 US
presiden tial elections conducted by McCombs and Shaw (1972). They
conclud ed th at th e media appeared to have had a significant influence on
voters’ judgements of which the major issues of the campaign were and that
there was a strong relationship between the emphasis given by the media to
differen t campaign issues and the salience which voters applied to them.
Early research defined agenda-setting in terms of the influence of the media
agend a on the public one. However, the process by which issues emer ge seems
to be b oth iterative and interactive and is by no means always initiated by the
media. MacKuen and Coombs (1981) were the first to find evidence (although
more suggestive than conclusive) that the press was ``the pr imary causal a gent’’
(p. 23) while there was some minor feedback from the public to the press. Their
conclusion that the influence of newspapers was more persistent than that of
television is supported by work of Benton and Frazier (1976), Clarke and Fredin
(1978), Asp (1983) and Allen and Izcaray (1988). A systematic definition of
agend a-setting as a process, in which the media, public and political agendas
interact with one another and with their mutual ext ernal environment, was
prop osed by Manheim and Albritton (1984) and elaborated by Man heim (1987).
Each agenda has its own internal dynamics an d each one is linked to the others
by informational, behavioural and institutional link s and could be represented
in diagrammatic form (Figure 1) (Rogers and Dearing, 1988).
Agenda -setting and campaign strategy
Becker (1977), after studying the 1972 U S Presidential campaign, concluded
that each party strategist tried to shift the attention of the campaign to the
issues which were most favourable to his candidate. Campbell et al. (196 6)
argu ed that parties should try to increase the salience of advantageous
``valence’’ issues (those propositions or beliefs which are positiv ely or
negativ ely valued by all voters) (see also Bowers, 1977; Brosius and
Kepp linger, 1992).
Who sets the
agenda?
1119
O’Keef e (1975) suggests that candidates should at least ``pay lip service’’ (p . 146)
to the issues wh ich concern the electorate but , in turn, they should also
attempt to direct voter attention to the issues they consider a s important. This
should be seen in conjunction to the concept o f issu e ownership by parties
(Ansolabeh ere and Iyen gar, 1994). Voters have prior beliefs about a party’s or
candid ate’s ability to deal with areas such as the economy, defence or welfar e.
The only communication tools over which the candidates can exert complete
control, as far as the convey ed message is concerned, are political advertising,
press conferences, speeches and policy statements. O’Keefe and Atwood (1981)
define the purposes of political advertising as twofold: to boost the morale of
campaig n workers; and to make available candidate and party information
that have not been filtered throu gh the news media. Roberts and McCombs
(1994), after studying the televised political advertisements that were
broad cast during the 1990 Texas gubernatorial campaign, concluded that
political advertising had a sign ificant agenda-setting influence in both the
press and television news agendas. Ansola behere et al. (1991) concluded that
paid political advertisements were more important in less newsworthy
campaig ns.
Norp oth and Buchanan (1992) argue that candidates should never follow a
strategy of ``trespassing’’ on the issue territory of another party. They suggest
that such a strategy runs the risk of raising issues and images that the public
has favourably associated with the opposing part y. Evidence for this in
practice is provided by Petrocik (1996) who reports that, during the 1980
Presidentia l election, the agendas of the New York Times and the two
candida tes were very differen t. Carter and Reagan tailor ed their activities and
Figure 1.
Three main components
of the agenda-setting
process
European
Journal of
Marketing
35,9/10
1120
their rep orts to match issues they each owned rather than adapt them to the
press agenda.
Kleinnijen huis and De Ridder (1998) tested two issue-voting theories:
(1) the issue ownership theory which argues that a party will be preferred
by the voters whenever the issues it owns dominate the media agenda;
and
(2) the issue position theory which argues that the voters’ ch oice will be the
result of a comparison between their own issue position and the current
positions of the parties.
Both theories offered explanations of the electoral outcomes in Germany and
The Neth erland s in 1994.
On the basis of the literature, one might expect to find that party agendas do
not necessarily correlate with the apparent press agenda. Clearly, parties and
candid ates seek to influence the media agend a in their fav our . However, the
literatur e does not offer clear guidance about how they might adapt their
strategy as the campaign unfolds, particularly for those falling beh ind.
Agenda -setting and political marketing
Ther e are only limited references in the political marketing literature to agenda-
setting and the press. O’Shaughnessy (1990) and Newman (1994, 1999) make no
specific reference to the agenda-setting role of the press in spite of the latter
consider ing newsp apers as ``one of the key power brokers in the political
process’’ (Newman, 1994, p. 30). He argues that its role during election
campaig ns has been enhanced by the emergence of investigative journalism. In
her assessment of the 1987 UK General Election campaign, Scammell (1995)
mention s agen da-sett ing in the particular context of political parties trying to
influence the media agenda. Maarek (1995) describes agenda-setting as the
influence of the media agenda on the public agenda, which was the original
concept, and refers to just three studies on t he subject, the latest being 1981.
Kava nagh (1995) noted the interaction and influence between press and
television and the way in which stories which app ear in o ne mediu m may
influence the agenda of the other during an election campaign. The apparent
failure of press conferences to influence the political agenda in the 1992 UK
Gener al Election was discussed by Kavanagh and Gosschalk (1995). Harris et
al. (1999) studied press coverage leading up to the 1997 UK General E lection,
which was dominated by a number of key issues, notably ``sleaze’’. They
conclud ed that these issues were not necessarily initiated by news media, but
that, once established, they acquired a momentum over which the p olitical
parties, particularly the Conservatives, had little influence.
Greece and the Greek electorate
Greece is a parliamentary democracy with a population of about 10.5 million.
The president is elected for a five-year term and there is a unicameral
parliam ent with some 300 members, elected ever y four years. Greece became a
Who sets the
agenda?
1121
member of the European Economic Community in 1981. Elections to the
Europ ean Parliament took place in 1981, 1984, 1989 and 1994, as in all member
states. There are seven parties ± PASOK, the ruling party (socialist), New
Democr acy (ND) (conservative), KKE (commu nist), Democratic Social
Movem ent (left-wing), Coalition (left-wing), the Liberals, and Political Spring (a
break away from New Democracy; conservative). At the beginning of the 1990s
PASOK and N ew Democracy acco unted for 80 per cent of the votes cast in
genera l elections. This proportion has fallen, but Greek politics are still very
much a duo poly.
These two parties have dominated the political scene of the country since the
fall of the military regime and the restoration of democracy in Greece in 1974.
PASOK has been in power since 19 93, latterly led by the ``modernisers’’ in the
party . As in other European countries, the basic div iding line in the Greek
party system a fter the second World War was the differentiation between Left
and Right in politics. This is still the basis of political allegiance for most
Greeks, though it is weakening with increasing evidence of growth in the centr e
and on the right of the spectru m (Vernadakis, 1999).
The study
The study consists of a co ntent analysis (Krippendorf, 1980) of newsp aper
articles which appeared in six Greek newspapers during the period 3 May until
12 June 1999. This period covers the election campaign, which started on
Mond ay, 10 May, the ``official’’ agreed starting date of the campaign period for
all parties. Press coverage of the week before that was included, as the parties
had already begun their campaigns before the ``official’’ date. The actual
election day was Sunday, 13 June.
The newspapers chosen for analysis were two mo rnin g papers Kathimerini
and To Vima and four afternoon newspapers, Ta Nea, E lef therotypia,
Elefth eros Typos and Ethnos. These papers had been selected for the analysis
on their respective size of circulation and spread of readers. The two morning
newspap ers make up 80 per cent of the copies of m orn ing newspaper sales and
the four afternoon ones make up 70 per cent of the copies of afternoon
newspap ers sold during the same period. In aggregate, according to the
Athenian Daily Newspapers Owners’ Union (EIIEA), the six accounted for
more than 72 per cent of all newspapers sold in Greece in May and June 1999.
This meant that the news content of the p ress in the period was covered in
some depth and allowed the agenda-setting role of th e press in Greece to be
assessed. The issue categories used to classify the content are the ones used by
the Public Opinion Poll Institute (V-PRC) in its studies of political opinion in
Greece (Mendrinou, 1999). The one alteration that was made was to merge the
categories of ``Criminality’’ and ``Illegal immigrants’’, as these issues had
become closely associated in the context of this election.
The press content analysed was taken from the main ``body’’ of each
newspap er ± those pages directly concerned with political and social content.
Intern ational pages, fashion pages, socialite pages, supplements, magazines
European
Journal of
Marketing
35,9/10
1122
and financial pages were excluded from the analysis, unless reference to
articles in these pages had been made in the front-page coverage of the
newspap er. In that case, these specific articles were included. Sports pages
were completely excluded. Newspapers issues were collected daily, apart from
the Sunday editions. Coverage was measured in terms of square centimetres as
well as the number of articles.
The campaign agenda of each party was traced throug h content analysis of
party manifestos for the 1999 elections and changes in the party agenda of the
two major parties (ND an d PASOK) was monitored th roug h the press releases
issued by them during the election camp aign. Walters et al. (1996) argue that
press releases are an important element of a party’s integrated political
strategy , amplifying campaign themes and images and generating media
coverag e. In analysing these we u sed the number of issues rather than
measur ing space. New Democracy, as the main opposition party, had formed
shadow ministries, which also issued press releases through the party ’s press
office. PAS OK ’s press office issued p rimar ily party press releases. In order to
make the two b odies of press releases comparable, the body of the government
annou ncements issued by the Prime Minister or Ministers during the given
period , were incorporated in the body of PASOK’s press releases. The public
agend a was identified from the results of a survey conducted by the V-PRC
Institut e in April 1999.
Wheth er the article appeared on the front page or inside pages was record ed.
A distinctio n was also m ade between editorial comment, comments from
prom inent journalists and explicit news coverage. It was expected that each
newspap er would put the stories which it considered as importan t on the front
page and these stories were most likely to be the subject of editorial or
prom inent journalists’ commentary (Shaw, 1 977; Iyengar and K inder, 1987;
Watt et al., 1993). The significant role of editorials in influencing people in all
ideological groups has also been studied (MacKuen and Coombs, 1981; Entman,
1989). The amount of space given to an issue in an editorial is an indicator of
the importance the editor places on the topic an d thus of the importance the
reader should attach to it.
The method in context
The period studied (3 May to 12 Ju ne 1999 ± 40 days) is similar t o those in other
election agenda-setting research. Siune and Borre (1975) studied a period of
three weeks, Asp (1983) a month-long perio d, Sem etko et al. (1991) a period of
24 days, and Harris et al. (1999) a period of 45 days.
Content analy sis is the dominant approach in agenda-setting research.
Semetk o et al. (1991) present the most comprehensive study of the influence of
the media in campaign agendas both in the UK and the USA and state that
althou gh content analysis does not shed any light into the behind-the-scenes
forces and relations which result in producing a daily newspaper, it ``can
docum ent what the media have covered’’ (p. 183). According to Phillips (1992)
content analysis measur es actual media behaviour. Dearing and Rogers (1996),
Who sets the
agenda?
1123
in an overview of the methodology used in agenda-setting research, mention
only content analysis as the method to determine the med ia agenda. They
empha sise that content analysis in typical agenda-setting research is concerned
with the aggregate number of news stories a nd not their exact content.
The majority of studies including measu rement of the press agenda have
used the total numb er of news stories about an issue (McCombs and Shaw,
1972; Tidmarch et al., 1 984; Semetko et al., 1991 ; McCombs et al., 199 7). Fewer
studies (Stone and McCombs, 1981; Salwen, 1988) have used the total amount of
space devoted to an issue (usually in column inches), while others have used
just the number of front-page stories (Wanta and Wu, 1992; Zhu, 1992). There is
still a debate on which is a better measure of the press agenda. MacKuen and
Coombs (1981) argue that the simple tot al of articles which are devoted to an
issue reflects the editor’s judgement of its salience (see also Mazu r, 1989).
Phillips (1992) and Watt et al. (1993) favour measurement in column inches
because it gives an idea of the level of authority and impact th e article conveys
on the reader and also shows how much space the newspaper is prepared to
devote to the issue when a ll related competitive coverage is taken into account.
Stone and McCombs (1981) compared the total number of stories and the total
numb er of colum n inches devoted to the issues o f the ag endas. This yielded a
correlation of +0.90 across all issues. They concluded that a simple count of
stories might be sufficient for the measurement of the media agenda. We
measured both space in square centimet res (equivalent to column inches) and
the number of articles.
We follow Phillips (1992) in including both headline and photographs as well
as text in the measurement of press coverage, as this reflects the impa ct o f an
article on the r eader. The same method was applied in the measurement of
parties’ manifestos. The press releases consisted of text only with out
photo graphs. Though photogr aphs have been the subject of separate conten t
analysis and found to have agenda-setting influence (Wanta, 1986) they were
not separately analysed by us. Cartoons were n ot included in our study though
they have also been the subject of research during election campaigns
(Seymou r-Ure, 1986).
The different agendas
The public agenda was taken from the results of a nation wide public opinion
poll conducted by V-PRC Institute in April 1999 (Mendrinou, 1999), just before
the start of the campaign. Table I shows the percentage of the sample reporting
an issue as their principal concern.
The press coverage of all six newspapers on every issue was analysed for
each category of issue. The totals are shown in Table II. The European election
coverag e is shown separately b y way of compariso n, though it is not one of the
ten identified issue categories. Other issues cover the full range of other stories
of public interest (e.g. personalities, accidents etc.).
From the results of all press coverage categories, it appears that the press as
a whole has its own a genda, promoted through the level of coverage of issues.
European
Journal of
Marketing
35,9/10
1124
Table I.
The initial public
agenda
Issues Percentage
International issues and foreign policy 28.56
Economy 27.81
Unemployment 25.14
Crisis of institutions and values 5.10
Criminality and illegal immigrants 4.02
Education 3.00
Health and social welfare 2.43
Other social iss ues (e.g. drugs) 1.78
Environment and life quality 1.64
Other issues 0.51
Source: Mendrinou (1999, p. 55)
Issues
Front page
cm
2
Editorial
cm
2
Comment
cm
2
News cover
cm
2
International issues and foreign
policy
50,299.15
(1)
16,446.75
(1)
156,332.1
(1)
913,546.9
(1)
Other issues 11,811.15
(2)
1,934.7
(3)
8,942.05
(2)
232,687.9
(2)
Health and social we lfare 10,515.15
(3)
2,206.35
(2)
6,765.6
(3)
125,464.25
(4)
Criminality and illegal immigr ants 10,178.8
(4)
1,603.9
(5)
6,674.4
(4)
144,515.5
(3)
Economy 7,736.25
(5)
1,443.85
(6)
2,047.8
(6)
70,533.45
(5)
Education 3,032.55
(6)
865.85
(8)
1,081.5
(10)
68,904
(6)
Environment and life quality 2,366.5
(7)
1,023.25
(7)
1,087.95
(9)
67,018.7
(7)
Unemployment 2,305
(8)
541.8
(9)
1,789.6
(7)
24,006.05
(10)
Other social is sues (e.g. drugs) 2,278.2
(9)
376.75
(10)
1,152.7
(8)
34,830.1
(9)
Crisis of institutions and values 2,019.4
(10)
1,648.55
(4)
3,940.5
(5)
56,939.8
(8)
European elections 17,148.3 6,653.25 61,037.3 439,933.35
Note: Column rankings in bracke ts
Table II.
The aggregate press
agenda
Who sets the
agenda?
1125
One might also argue that, regardless of political affiliation s and ideological
preferen ces, newspapers tend to have similar journalistic norms and criteria of
newswort hiness. We are satisfied t hat the aggregate newspaper agenda is
represen tative of the six newspapers’ coverage of issues and can be used in the
research.
The content-analysis of the party manifestos produced the space coverage
shown in Table III. The parties’ aggregate agenda was not calculated as the
objective was to draw conclusions for each specific party, and especially for the
two major ones, and not for the parties as a whole.
The press releases for the two major parties during the election campaign
were analysed using the same ten categories and are shown in Table IV.
However , given the different nature of press releases, just the number of clear
references to each category was recorded. It is interesting to n ote that two-
thirds of the ruling party (PASOK) press relea ses made no explicit reference to
any of the identified categories.
The agendas of the mainstream parties wer e cor related in order to
determin e whether they, despite their ideological differences, assigned similar
import ance to issues (Table V). For this analysis, Spearman’s Rho rank order
Issues
PASOK
cm
2
ND
cm
2
KKE
cm
2
DSM
cm
2
Coalition
cm
2
Pol
Spring
cm
2
Liberal
cm
2
Economy 7,527
(1)
1,734
(1)
2,549
(1)
361
(2)
110
(4)
230
(2)
180
International issues and
foreign policy
4,761
(2)
198
(4)
2,207
(2)
725
(1)
1,953
(1)
2,450
(1)
144
(2)
Health and social welfar e 3,933
(3)
114
(6)
160
(3)
128.5
(3)
170
(2)
80
(7)
135
(3)
Other issues 2,964
(4)
1,120
(3)
0
(8=)
0
(10)
0
(9=)
0
(8=)
0
(5=)
Other social is sues 1,331
(5)
150
(5)
155
(4)
66
(4)
112
(3)
130
(4)
0
(5=)
Environment and life
quality
1,083
(6)
110
(7)
0
(8=)
30
(7)
20
(7)
0
(8=)
0
(5=)
Unemployment 622
(7)
76
(8)
77
(7)
6
(9)
70
(5)
90
(6)
0
(5=)
Crisis of institutions and
values
533
(8)
34
(10)
96
(5)
38
(6)
55
(6)
170
(3)
180
(1)
Education 328
(9)
51
(9)
20
(6)
54
(5)
10
(8)
0
(8=)
0
(5=)
Criminality and illegal
immigrants
239
(10)
1,731
(2)
0
(8=)
25
(8)
0
(9=)
110
(5)
75
(4)
Note: Column ranks in brackets
Table III.
The party agendas at
the beginning of the
campaign
European
Journal of
Marketing
35,9/10
1126
correlatio n coefficient was chosen. The 5 per cent and 1 per cent significance
levels for sa mples of ten pairs are 0.564, and 0.746 respectively (Siegel, 1956).
These are indicat ed in the tables by a single asterisk and a double asterisk
respectively .
Five correlations are significant at the 5 per cent level (*) and four are at the 1
per cent level (8). Twelve correlations are not significant at the 5 per cent level.
These results suggest that there are two blocks of agendas in the Greek
political scene. In the one block, there is ``New Democracy’’ (the conservativ e
party), whose agenda did not correlate significantly with the a gend a of any
other party, and in the other block we find PASOK (th e socialist party), KKE
Issues
New democracy
Number of press
release references on
subject (N = 153)
PASOK
Number of press
release references on
subject (N = 91)
International issues and foreign policy 24 (1) 10 (1)
Other issues 18 (2) 6 (3)
Economy 13 (3) 3 (4)
Crisis of institutions and values 12 (4) 1 (6=)
Criminality and illegal immigra nts 10 (5) 2 (5)
Unemployment 10 (6) 0 (8=)
Health and social we lfare 7 (7) 8 (2)
Environment and life quality 4 (8) 0 (8=)
Education 3 (9) 0 (8=)
Other social is sues 3 (10) 1 (6=)
European elections 49 60
Total 153 91
Note: Column ranks in brackets
Table IV.
Table V.
Correlations between
party agendas
PASOK ND KKE DSM Coalition
Political
Spring
The
Liberals
PASOK 0.491 0.693* 0.588* 0.663* 0.362 ±0.020
ND 0.166 0.164 0.012 0.301 ±0.145
KKE 0.865** 0.88** 0.739* 0.297
DSM 0.796** 0.607* 0.328
Coalition 0.597* 0.318
Political Spring 0.498
The Liberals
Who sets the
agenda?
1127
(the communist party), DSM and Coalition (left-wing parties) and ``Political
Sprin g’’ (right-wing party). The first four parties o f the second block have high
correlation s among their ag endas, wh ich indicates a stro ng relationship. This
may well be due to the fact that all four target more or less the same par t of the
electorate (centre-left and left-wing voters). The surprising result is that
``Political Spring’’, in spite of being a right-wing party (splinter from ``New
Democr acy’’), has an agenda which correlated highly with the agenda of the
three left-wing parties (KKE, DSM and Coalition). This by no means suggests
that the parties, whose agendas showed a high level of correlation, propose the
same policies. It means that these parties have identified the same issues and
rank them similarly.
Another interesting result is the moderately high correlation between the
New Democracy and PASOK agendas in spite of not r eachin g the 5 per cent
level of statistical significance (rho = 0.491). Taking into consideration that
PASOK ’s agenda showed th e highest correlation a mong ``left-wing’’ parties’
agend as with the one of New Democracy, this indicates that the two major
parties may have a quite similar view of which are the im port ant issues for
contemp orary Greek society. An alterna tive explanation is that they are
respon ding to one another as the campa ign unfolds.
Thou gh the election was for the European Parliament, clearly ``internal’’
issues (such as ``Health and social welfare’’, ``Crisis of institu tions and values’’
and ``Criminality and illegal immigrants’’) form a significant part of the agendas
of all the parties. The focus of the campaign on issues of ``internal’’ (national)
interest was particularly obvious in the case of the two m ajor part ies. One
explan ation is that the 1999 European election was a ``preliminary test’’ for the
Genera l Election in 2000. It is not unusual for European Elections to be fou ght
on domestic political lines in member states. The centrality of domestic or
interna l issues in the Greek E uropean Election campaign suggests that agenda-
setting is relatively unaffected by the type of elect ion.
The agg regat e press and the parties’ agendas were also correlated in order to
determin e the strength of the relationship between them (Table VI).
The vast majority of correlations were not significant at the 5 per cent level.
The results for the Liberals are interesting as they had only four issues in their
Table VI.
Front page Editorial Commentary News reporting
PASOK 0.467 0.394 0.430 0.358
ND 0.612* 0.261 0.467 0.612*
KKE 0.080 0.202 0.227 0.031
DSM 0.2 0.248 0.115 0.212
Coalition 0.067 0.176 0.213 ±0.049
Political Spring ±0.018 0.215 0.362 0.080
The Liberals 0.164 0.710* 0.608* 0.328
European
Journal of
Marketing
35,9/10
1128
agend a. New Democracy is the only other party with an apparent relationship
with th e press agenda, and then in the ``Front page’’ and ``News rep ortin g’’
rather than the ``Editorial’’ and ``Commen t’’ categories. These r esults do not
mean that ``N ew Democracy’’ r eceived favourable coverage for its views; they
simply mean t hat the party ranked the issues similarly to th e newspapers in
aggr egate.
The mo st important conclusion to be drawn from these results is that
althou gh the parties and the press have distinct agendas, the press and political
parties do not have a common view of the relative importance of the issues in
Greek society.
Similarly , rank order correlation s were calculated between the aggregate
press agenda and the public agenda (Table VII).
These results indicate that there is no strong relationship between the public
agend a an d the aggregate press agenda. The correlations are very low for all
four categories of news coverage. This is particularly interesting given that the
measur ement of pub lic opinion preceded the period of our measurement of the
aggr egate press agenda. By contrast, the correlations between the party
agend as and public opinion show some significant relationships in a number of
cases and, generally, apparently a higher level of association, though the low
figures for the two leading parties are striking (Table VIII). Indeed, the o verall
results suggest t hat the public and party agendas are distinctly different.
From a political marketing point of view, it seems that most Greek parties
(especially the two major ones), in sp ite of doing their market research (focus
grou ps, opinion polls, etc.) choose to attach different relative importance to
Table VII.
Public agenda
Front page 0.055
Editorial 0.164
Commentary 0.212
News reporting 0.055
Table VIII.
Public agenda
PASOK 0.164
New Democracy 0.079
KKE 0.620*
DSM 0.479
Coalition 0.450
Political Spring 0.779**
The Liberals 0.403
Who sets the
agenda?
1129
issues than electors do. This conclusion is suppo rted by a recent study from the
Univer sity of Piraeus (Roubanis, 2000) using the Delphi Met hod with political
correspo ndents who work in t he media. The correspondents reportedly believe
that the parties do not resp ond to contemporary needs satisfactorily, as they
have neither identified the major issues of the coun try nor formed policies
which will successfully solve them.
The 1999 Elections for the European Parliament showed interesting cha nges
in the allegian ce of th e elector ate (Table IX) with smaller parties gaining at the
expense of the two leading ones. Whilst this might be attributed partly to the
lower significance that voters in the EU attribute to European as opposed to
nationa l elections, disaffection with pa rties and the electoral system appears to
be increasin g. The proportion of blank an d spoilt ballot papers (Figure 2) and
the level of abstention is r ising for both national and European electio ns,
thoug h it is still higher for the latter (Figu re 3). The share of ND and PASOK in
the actual election fell to a new low of just under 69 per cent .
The apparent lack of a strong relationship among the public agen da and the
two major parties’ agendas shown in this study may provide some explanation
for this decline in popularity of the two major parties and evidence of a more
genera l disenchantment with elections (Figure 4). Small parties’ agendas
showed much higher correlations with the public agenda, which may account
for their growing public support.
Table IX.
Trends in the political
identification of the
electorate (1996-1999)
Political identification
1996 (general elections)
(%)
1999 (European elections)
(%)
With PASOK 41.6 34.3
With New Democracy 37.5 32.0
With small parties 10.8 14.6
With no party at all 14.9 21.8
Source: Nikolakopoulos (1999)
Figure 2.
The evolution of ``blank’’
and ``sp oilt’’ ballot
papers (1993-1999)
European
Journal of
Marketing
35,9/10
1130
Discu ssion
The findings of previous studies on the relationships among the press, parties’
and public agendas are somewhat confused. In respect of the relationship
between the press agenda and the parties’ agendas, our findings are similar to
those of Tichenor (1982), who also found great disparities amon g them du ring
the 1976 Presidential Election. On the contrary, Dalton et al. (1998), after
study ing the 1992 Presidential Election in the USA, found significant
agreem ent between the press and candidates’ a gend as. Our results are
consistent with the camp aign strategy literature which imp lies that parties
should focus on the issues they ``own’’ and not be drawn on opponents’
territor y.
Analysin g the data on the relationship between the press agenda and the
public agenda is con cerned, the finding s of this study are again in discord with
those of Dalton et al. (1998) who found that the match between the press and
public agendas w as ``striking’’ (p. 474).
Figure 3.
Evolution of vote rs’
abstention
Figure 4.
Evolution of the two
major parties’ aggregate
percentage vote
Who sets the
agenda?
1131
The findings support the issue ownership theory (Kleinnijenhuis and De
Ridder , 1998). This theory states that a party will be preferred by the voters
when the issues it owns dominate the media a genda. ``New Democracy’s’’
agend a showed a strong er relationship with the pr ess agenda than the other
parties’ agendas did, which may have facilitated its electoral victory, though
we offer this conclusion with caution given that we did not study broadcast
media coverag e. The results do not seem to support Kleinnijenhuis and De
Ridder ’s issue position theory which states that the voters will prefer the party
whose issue positions r eflect their own. ``New Democracy’’ won the elections
(albeit with a reduced vote) in spite of its agenda having the lowest correlation
with the priorities of the public agenda.
The Greek press seems to play a role similar to the one played by the US
Press (according to Semetko et al., 1991) as far as the formation of the campaign
agend a is concerned. The press had great discretion in emphasising or
downp laying issues, although this ability seemed to become limited during the
campaig n. This role has been termed by Semetko et al. (1991) as ``agenda
shapin g’’. Missika and Bregman (1987) suggest that the formation of the
campaig n agenda is achieved th rough interaction between the media and the
politicians. Voters are obliged to accept the outcome of this interaction
althoug h they can react to it; what they cannot do is to act proactively in order
to influence it. As the agenda does not reflect the public worries, the electorate
is disenchanted and votes for smaller parties, abstains or casts blank or spoilt
ballots. This is a good description of the ob servable trends in the Greek 1999
Europ ean Elections. Similar p henomena may be observed elsewhere in Europe,
thoug h manifestations of it vary according to the particular form of electoral
system.
Studies of this kind have limitat ions. First, whilst the issues are identified,
the actual positions of the differ ent parties and the press are not. We have noted
that broadcast media w ere not studied. Also for the purposes of this analysis
we treated press coverage as a w hole, aggregating across the six newspapers,
as we were interested in the ov erall press agenda rather than identifying
differing perspect ives between them However, the results do appear to tie in
with results from other analyses of the elections under consideration.
Conc lusions
The result that the press agenda in ter ms of the relative priority assigned to
issues, whether in editorial, other comment or core news coverage, differs from
both the party agendas and the public one is a striking one. H owever, we do not
believe necessarily that the press delib erately sets out to construct an agenda to
influence elections, though papers may hav e strong party allegiances. There
does appear in a number of national press settings to be some consensus about
what makes newsworthy stories at a ny particular time during an election.
These stories and themes are som etimes initiated by the parties themselves,
sometimes by what are considered ``gaffes’’, and sometimes by ex ternal events.
What does appear clear is that parties, even if they manage to initiate cover age
European
Journal of
Marketing
35,9/10
1132
of a specific issue, do not subsequently manage to adapt the overall press
agend a to their specific priorities. The difference between the public and the
press agendas is not entirely consistent with studies in other national settings,
but it may be evidence to suggest that the public does not necessarily respond
to the press agenda and, that if they do converge, the mov ement may be from
either side.
We tentatively link the divergence between the parties’ and the public
agend as to making the electorate feel isolated and ignored by its political
leaders and thus become dissatisfied with the p olitical system and the parties
(first the agendas diverge, then comes the disenchantment); this is a matter of
concern in many democracies. T here are however, other, though not necessarily
mutu ally exclusive, explanation s. First, the need to keep party a ctivists content
plays a significant role in the constr uction of political manifestos and other
messages. Second, the need to construct distinct market positions may lead to
differen t emphases on issues and, at least in the short term, a party’s
positionin g strategy has to be reasonably consisten t with voters’ perceptions.
Finally, the campaign strategy literature emphasis on issue ow nersh ip and
parties sticking to their own issue territory may actually inhibit agenda
adapt ation. It offers little guidance to parties or candidates who fall behind in
campaig ns. We now have good models of the agenda-setting process. What
appear s to be needed at this point are theories on which to base campaign
strategies which are capable of adapting and respon ding as events unfold.
References
Allen, R.L. and Izcaray, F.C. (1988), ``Nominal a genda diversity in a media-rich, less-developed
society’’, Communi cation Research, Vol. 15 No. 1, pp. 29-50.
Ansolabehere, S . and Iyengar, S. (1994), ``Riding the wave and claiming ownership over issues:
the joint effects of advertising and news coverage in campaigns’’, Public Opinion Quarterly,
Vol. 58, pp. 335-57.
Ansolabehere, S., Behr, R. and Iyengar, S. (1991), ``Mass media and elections : an overview’’,
American Politics Quarterly, Vol. 19 No. 1, pp. 109-39.
Asp, K. (1983), ``The struggle for the agenda: party agenda, media agenda and voter agenda in
the 1979 Swedish election campaign’’, Communication Re search, Vol. 10 No. 3, pp. 333-55.
Becker, L.B. (1977), ``The impact of issue saliences’’, in Shaw, D.L. and McCombs, M.E. (Eds), The
Emergence of American Political Issues: The Agenda-Setting Function of the Press, West
Publishing Co, St Paul, MN, pp. 121-31.
Benton, M. and Frazier, P.J. (1976), ``The agenda-setting function of the mass media at three levels
of `information-holding’’’, Communication Research, Vol. 3, pp. 261-74.
Bowers, T.A. (1977), ``Candidate advertising: the agenda is the message’’, in Shaw, D.L. and
McCombs, M.E. (Eds), The Emergence of American Political Issues: The Agenda-Setting
Function of the Press, We st Publishing Co, St Paul, MN, pp. 5 3-67.
Brosius, H.-B. and Kepplinger, H.M. (1992), ``Beyond agenda-setting: the influence of partsansh ip
and television reporting on the electora te’s voting intentions’’, Journalism Quarterly,
Vol. 69 No. 4, pp. 893 -901.
Campbell, A., Converse, P.E., Miller, W.E. and Stokes, D.E. (1966), Elections and the Political
Order, John Wiley and Sons, New York, NY.
Who sets the
agenda?
1133
Clarke, P. and Fredin, E. (1978), ``Newspapers, television and political reasoning’’, Public Opinion
Quarterly, Vol. 42, pp. 143-60.
Cohen, B.C. (1963), The Press and Foreign Policy, Princeton University Press, Princeton, NJ.
Dalton, R.J., Beck, P.A., Huckfeldt, R. and Koetzle, W. (19 98), ``A test of media-centered agenda-
setting: newspaper content and public interests in a presidential election’’, Political
Communication, Vol. 15, pp. 463-81.
Dearing, J.W. and Rogers, E.M. (1996), Agenda-Setting, Sage, Thousand Oaks, CA.
Entman, R.M. (1989), ``How the media affect what people think: an information processing
approach’’, Journal of Politics, Vol. 51 No. 2, pp. 347-70.
Harris, P., Lock, A.R. and Roberts, J. (1999), ``A content analysis of press coverage of p olitical
issues during the 1997 UK general election campaign’’, in Newman, B.I. (Ed.), Handbook of
Political Marketing, Sage, Thousand Oaks, CA, pp. 539-58.
Iyengar, S. and Kinder, D.R. (1987), News That Matters: Television and A merican Politics,
University of Chicago Press, Chicago, IL.
Kavanagh, D. (1995), Election Campaigning: The New Marketing of Politics, Blackwell Publishers
Ltd, Oxford.
Kavanagh, D. and Gosschalk, B. (1995), ``Failing to set the agenda: the role of Election Press
Conferences in 1992’’, in Crewe, I. and Gosschalk, B. (Eds), Political Communications: The
General Election Campaign of 1992, Cambridge Univers ity Press, Cambridge, pp. 160-74.
Kleinnijenhuis, J. and De R idder, J.A. (1998), ``Issue news and electoral volatility’’, Euro pean
Journal of Political Research, Vol. 33, pp. 413-37.
Krippendorf, K. (1980), Content Analysis: An Introduction to its Methodology, S age, Newbury
Park, CA.
Lippman, W. (1922), Public Opinion, Harcourt Brace, New York, NY.
Maarek, P.J. (1995), Political Marketing and Communication, John Libbey and Company Ltd,
London.
MacKuen, M.B. and Coombs, S.L. (1981), More Than News : Media Power in Public Affairs, Sage,
Beverly Hills, CA.
Manheim, J.B. (1987), ``A model of agenda-dynamics’’, in McLaughlin, M.L. (Ed.), Communication
Yearbook 10, Sage, Newbury Park, CA, pp. 499-516.
Manheim, J.B. and Albritton, R.B. (1984), ``Changing national images: international public
relations and media agenda-setting’’, The American Political Science Review, Vol. 78 No. 3,
pp. 641-57.
Mazur, A. (1989), ``Commu nicating risk in the mass media’’, in Peck, D.L. (Ed.), Psychosocial
Effects of Hazardous Toxic Waste D isposal on Communities, Charles C. Thomas,
Springfield, IL, pp. 119-37.
McCombs, M.E. and Shaw, D.L. (1972), ``The a genda-setting function of mass media’’, Public
Opinion Quarterly, Vol. 36, pp. 176-87.
McCombs, M.E. and Shaw, D.L. (1977), ``The agenda-setting function of the press’’, in Shaw, D.L.
and McCombs, M.E. (Eds), The Emergence of American Political Issues: The Agenda-
Setting Fu nction of the Press, West Publishing Co, St Paul, MN, pp. 1-18.
McCombs, M.E., Llamas, J.P., Lopez-Es cobar,E. and Rey, F. (1997), ``Candidate images in Spanish
elections: second-level agenda-setting effects’’, Journalism and Mass Communication
Quarterly, Vol. 74 No. 4, pp. 703-17.
Mendrinou, M. (1999), ``Public opinion, political demands and `issu e publics’’’, in V-PRC Institute
(Ed.), Public Opinion in Greece: Surveys-Opinion Polls 1999-2000, ``Nea Synora’’ ± A.A.
Livanis Publications, Athens.
European
Journal of
Marketing
35,9/10
1134
Missika, J.-L. and Bregman, D. (1987), ``On framing the campaign: mass media roles in
negotiating the meaning of vote’’, European Journal of Communication, Vol. 2, pp. 289-309.
Newman, B.I. (1994), The Marketing of the President: Political Marketing as Campaign Strategy,
Sage, Thousand Oaks, CA.
Newman, B.I. (1999), The Mass Marketing of Politics: Democracy in an Age of Manufactured
Images, Sage, Thousand Oaks, CA.
Nikolakopoulos, I. (1999), ``Shrinking of the political identification of the voters’’, Ta Nea,
19-20 June, p. 10.
Norpoth, H. and Buchanan, B. (1992), ``Wanted: the education president: issue trespassing b y
political candidates’’, Public Opinion Quarterly, Vol. 56, pp. 87-99.
O’Keefe, G.J. (1975), ``Political campaign and mass communication research’’, in Chaffee, S.H.
(Ed.), Political Communication: Issues and Strategies fo r Research, Sage, Beverly Hills, CA,
pp. 129-64.
O’Keefe, G.J. and Atwood, L.E. (1981), ``Communication and election campaigns’’, in Nimmo, D.D.
and Sanders, K.R. (Eds), Handbook of Political Communication, Sage, Beverly Hills, CA,
pp. 329-57.
O’Shaughnessy, N.J. (1990), The P henomenon of Political Marketing, The Macmillan Press,
Basingstoke.
Petrocik, J.R. (1996), ``Issue ownersh ip in presidential elections, with a 1980 case study’’,
American Journal of Political Science, Vol. 40 No. 3, pp. 825-50.
Phillips, D. (1992), Evaluating Pres s Coverage: a Practical Guide to Measurement and Cost
Effectiveness, Kogan Page, London.
Roberts, M. and McCombs, M. (1994), ``Agenda setting and political advertising: origins of the
news agenda’’, Political Communication, Vol. 11, pp. 249-62.
Rogers, E.M. and Dearing, J.W. (1988), ``Agenda-setting research: where has it been? Where is it
going?’’, in Anderson, J.A. (Ed.), Communication Yearbook 11, Sage, Newbury Park, CA,
pp. 555-94.
Roubanis, G. (2000), ``What will determine the outcome of the election’’, Eleftherotypia, 8 February,
p. 5
Roubanis, T. and Karimali, A. (1999), ``The surprises of the ballot-box’’, Ethnos, 15 June, p. 23.
Salwen, M.B. (1988), ``Effect of accumulation of coverage on issue salienc e in agenda-setting’’,
Journalism Quarterly , Vol. 65, pp. 101-6.
Scammell, M. (1995), Designer Politics: How Elections Are Won, Mac millan Press, Basingstoke.
Semetko, H.A., Blumler, J.G., Gurevitch, M., Wea ver, D.H. with Barkin, S. and Wilhoit, G.C.
(1991), The Formation of Campaign Agendas: a Comparative Analysis of Party and Media
Roles in Recent American and Britis h Elections, Lawrence Erlbaum As sociates,
Hillsdale, NJ.
Seymour-Ure, C. (1986), ``Drawn and quar tered: the election in cartoons’’, in Crewe, I. and Harrop,
M. (Eds), Political Communications: Th e General Election Campaign of 1983, Cambridge
University Press, London, pp. 16 0-76.
Shaw, D.L. (1977), ``The press agenda in a community setting’’, in Shaw, D.L. and McCombs, M.E.
(Eds), The Emergence o f American Political Issues: The Agenda-Setting Function of the
Press, West Publishing Co, St Paul, MN, pp. 19-31.
Siegel, S. (1956), Nonparametric Statistics for the Behavioral Sciences, McGraw-Hill Book
Company, New York, NY.
Siune K. and Borre, O. (1975), ``Setting the agenda for a Danish election’’, Journal of
Communication, Vol. 25 No. 1, pp. 65-73.
Stagkos, A. (1999), ``What do the figures show’’, To Vima, 15 June, p. A5.
Who sets the
agenda?
1135
Stone, G.C. a nd McCombs, M.E. (1981), ``Tracing the time lag in agenda-setting’’, Journalism
Quarterly, Vol. 58, pp. 51-5.
Tichenor, P.J. (1982), ``Agenda-setting: media as political kingmakers?’’, Journalism Quarterly,
Vol. 59, pp. 488-90.
Tidmarch, C.M., Hyman, L.J. and Sorkin, J.E. (1984), ``Press issue agendas in the 19 82
congressional a nd gubernatorial election cam paigns’’, Journal of Politics, Vol. 46 No. 4,
pp. 1226-45.
Vernadakis, C. (1999), ``The party system 1995-1998’’, in V-PRC Institute (Ed.), Public Opinion in
Greece: Surveys-Opinion Polls 1999-2000, ``Nea Synora’’ ± A. A. Livanis Publications,
Athens, pp. 17-43.
Walters, T.N., Walters, L.M. and Gray, R. (1996), ``Agenda building in the 1992 presidential
campaign’’, Public Relations Review, Vol. 22 No. 1, pp. 9-24.
Wanta, W. (1986), The Agenda-Setting Effects of Dominant Photographs, paper presented at the
annual meeting of the Association for Education in Journalism and Mass Communication,
Norman, OK.
Wanta, W. and Wu, Y.-C. (1992), ``Interpersonal communic ation and the agenda-setting process’’,
Journalism Quarterly , Vol. 69 No. 4, pp. 847-55.
Watt, J.H., Mazza, M. and Snyder, L. (1993), ``Agenda-setting effects of television news coverage
and the effects dec ay curve’’, Communication Research, Vol. 20 No. 3, pp. 408-35.
Zhu, J.-H. (1992), ``Issue competition and attention distraction: a zero-sum theory of agenda-
setting’’, Journalism Quarterly, Vol. 69 No. 4, pp. 825-36.
