The National Press and Party Voting in the UK
| DOI | 10.1111/1467-9248.00313 |
| Author | Malcolm Brynin,Kenneth Newton |
| Date | 01 June 2001 |
| Published date | 01 June 2001 |
| Subject Matter | Article |
The National Press and Party Voting
in the UK
Kenneth Newton and Malcolm Brynin
University of Southampton and University of Essex
The difficulty with resolving the classic problem of whether newspapers influence voting patterns
is self-selection: readers select a paper to fit their politics, and newspapers select particular types
of readers. One way round this chicken-and-egg problem is to compare the voting behaviour of
individuals whose politics are reinforced by their paper, with those who are cross-pressured by
their paper, and to compare both with those who do not regularly read a paper. Using the British
Household Panel study to analyse voting patterns in 1992 and 1997, this study suggest that news-
papers have a statistically significant effect on voting, larger for Labour than Conservative sympa-
thizers, and larger for the 1992 than the 1997 election. The broader implications of these findings
for British politics and democracy are discussed.
Theory
As the mass media become more pervasive in modern life, so the debate about their
role in political life becomes more controversial. For some it is self-evident that
media influences are weak and can do no more than reinforce pre-existing tenden-
cies; others are equally convinced that they are a powerful political force in their
own right. Social scientists have generally favoured the minimalist – reinforcing
position for most of the post-war period, but there is now a greater inclination to
consider the possibility of media effects that are not only strong but also direct.
However, as Iyengar (1997, pp. 215–6) argues:
research into the effects of mass communication has come full circle.
Initial concern about the vulnerability of voters to propaganda campaigns
gave way to findings of ‘minimal consequences’. Research was rejuvenated
by more limited conceptions of media influence, as manifested by the
learning, agenda-setting, and priming paradigms. As these paradigms
have matured, discussions of ‘massive consequences’ have been revived.
The mass media’s agenda-setting capacity has been accepted for some time now –
the ability to influence not what people think, but what they think about. Other
indirect influences of priming and framing roles were then added – the idea that
the media can prime people to react in one way rather than another, or can present
(frame) issues in a way that leads people to think about them in one way rather
than another. Most recently, there has also been an accumulation of findings
suggesting that the modern mass media can also have a strong and direct effect not
just on how people think about politics but what they think.1
Talk of ‘massive consequences’ is less common in Britain, but recent research
does suggest that the mass media can have a direct and significant effect on mass
political attitudes and behaviour, as well as a possible agenda-setting, priming, and
© Political Studies Association, 2001.
Published by Blackwell Publishers, 108 Cowley Road, Oxford OX4 1JF, UK and 350 Main Street, Malden, MA 02148, USA
266
K E N N E T H N E W T O N , M A L C O L M B R Y N I N
framing role.2 The debate about media effects, however, remains highly contro-
versial, and claims of almost every kind are contested.3
The controversy is difficult to resolve because it is exceedingly difficult to isolate
and measure the impact of the mass media. In the first place, there are different
forms of media – television, radio, newspapers, broadsheets, tabloids – each with
its own set of messages and effects, and each possibly reinforcing or neutralizing
the others. Media consumption is also closely linked to a range of social, economic,
and political variables that are themselves associated with political attitudes and
behaviour. Education, for example, is closely related to patterns of media consump-
tion, on the one hand, and political attitudes and voting behaviour, on the other,
making it exceedingly difficult to disentangle the separate media and education
effects. The same may well be true of variables such as class, occupation, income,
age, and gender.
The most difficult problem concerns the close cause-and-effect relations between
audiences and the mass media. Audiences select the media they attend to, just as
the media target different audiences and can do so with precision because, more
than most producers of consumer products, they benefit from elaborate market
research methods to help them identify and appeal to their consumers. They are
the ‘fruit-flies’ of modern consumerism; they sell to millions of people daily, they
can experiment with their products with comparative ease and speed, and they
work closely with the opinion polling industry. The commercial media, it is said,
are bound by the golden chains of the market; and no less than producers of wash-
ing powder they must give their customers what they demand, and are able and
willing to pay for. To this extent they are obliged by market forces to reflect, to
echo, and to follow mass opinion, rather than mould, influence, or change it.
There is plenty of anecdotal evidence showing that the mass media are prisoners of
the laws of market supply and demand. For example, the Sun switched its support
from Conservative to Labour in the 1997 election campaign, perhaps because it
knew that many of its readers were going to vote Labour anyway and did not want
to alienate its market (Seymour-Ure, 1997; Curtice, 1998). The normally staunch
Conservative Financial Times cautiously stated the paper’s wish not to see a
Conservative victory in 1992, and immediately suffered criticism and a dip in sales.
In a competitive market, where there is sensitive adjustment of supply and demand,
it is likely that most causal influences will run from consumers to producers, but in
any case, causes and effects running in either direction will be exceedingly difficult
to pin down.
All this makes research into media effects difficult – but not impossible. In fact,
by most national standards Britain is a rather good place for research. The press is
highly centralized, and most people read a national daily or Sunday paper. The
1992 British Household Panel Study (BHPS) shows that that 69 percent of the
sample regularly read a national daily newspaper, a figure very close to the 66 per-
cent found in 1996 by the British Social Attitudes Survey (Newton, 1997, pp. 154).
As a result, the circulation of the national daily and Sunday papers in Britain is
high by most western standards. Of the eleven daily and Sunday papers with the
largest sales in west Europe, nine are British. In round figures the eleven British
national dailies of 1992 (which include the Financial Times) sold 13.5 million copies
T H E N A T I O N A L P R E S S A N D P A R T Y V O T I N G I N T H E U K
267
a day, and had an estimated total readership of over 36 million. Daily sales of the
five largest national dailies (Sun, Mirror, Mail, Express, and Telegraph) are 10.7 million,
and readership exceeds 28 million (Harrop and Scammell, 1992; Scammell and
Harrop, 1997). This provides research with large numbers to sample.
In addition, the British press is highly partisan by most western standards; most
papers have clear party attachments, and most nail their political colours to their
mastheads. They do this in part because the electronic media – radio and TV – are
bound by law to present fair, impartial, and balanced political news, and to avoid
all forms of political advertising and editorializing. British newspapers take up clear
party political positions to distinguish themselves and their market from those of
the neutral electronic media. This also means that the partisan messages of the
newspapers do not have to compete with the electronic media – in this respect they
have the field to themselves. Large national markets encourage the papers to carve
out a particular niche for themselves, which means targeting particular social, eco-
nomic and political groups. Newspapers are usually less partisan in more localized
markets, where they try to maximize sales by appealing to a broad cross-section of
the community and avoid strong views on controversial matters (Dalton et al., 1998).
In sum, Britain has a national newspaper market with a relatively small number
of daily and Sunday titles, most with clear and strong party political positions, read
by large numbers of people, and without partisan competition from television
and radio. This combination of characteristics may be interpreted in two entirely
different ways – as so often in media research. Either the national newspaper
market ensures that consumers get what they want, in which case newspapers will
have to adapt themselves to the politics of their readers, and the press will have
little political influence of its own. Or the political influences (if there are any) of
strongly partisan papers with large national markets will show up more clearly in
Britain than in most other western nations.
The hypothesis to be investigated here is not that newspapers are the major
influence on party voting or even among the most important influences. Rather
the main proposition to be tested is whether the national press in Britain has a
statistically significant association with voting patterns, and if so, whether these
associations are independent of control variables that account for the attitudes and
values that readers bring to their paper in the first place.
It terms of election outcomes the national press might make no more...
To continue reading
Request your trial