Politicians frequently worry about the effect on the voters of newspaper editorials endorsing rival political parties during a general election campaign. Harold Wilson, for example, was said to be obsessed by the press coverage of his government and the likely electoral consequences of it. Similarly, the current prominence of spin-doctors and the growth of media advisers in government demonstrate the importance of having favourable media coverage in the modern never-ending election campaign. We know which newspapers endorsed which parties at the last election, although the extent to which they covered the campaign and the enthusiasm of these endorsements varied widely between them. To different degrees the Guardian, the Independent, the Financial Times, the Mirror, the Sun and the Daily Star endorsed New Labour in 1997. Equally and again with varying levels of enthusiasm the Daily Telegraph, the Express and the Daily Mail endorsed the Tories. The Times sat on the fence.
To what extent did readers follow these endorsements and vote for the parties in question? In terms of the strength of endorsement the two polar opposites were the Mirror and Telegraph. If editorials have an effect we would see pretty much all the Telegraph readers voting Tory and a similar proportion of Mirror readers voting Labour.
To be fair, if we do see an association between voting and editorial endorsements, we cannot be sure if it is because readers follow the advice of editors or because they choose newspapers which agree with their political preferences. Research suggests that at least some of the influence runs from editors to readers, so it is interesting to see how closely readers followed the endorsements of their papers.
The chart shows the relationship between voting behaviour and newspaper readership at the 1997 general election, using the British election study survey, which was conducted immediately after polling day. The newspapers are ranked in order of support for Labour. It can be seen from these rankings that, sure enough, the Mirror is at the top and the Telegraph at the bottom. But while the Mirror had the highest proportion of readers supporting Labour, it was not true that Mirror readers were the least likely to vote Conservative. This distinction went to Guardian readers. The election study, which consisted of about 2,500 randomly selected individuals, contained not a single Guardian-reading Tory. Guardian readers were more likely to turn out and vote than the readers of any other newspaper.
The highest levels of Labour support show among readers of the Mirror, the Guardian and the Daily Star. They were clustered together as a group with around 60% or more of their readers voting for Labour. A second cluster of readers included the Independent and the Sun with roughly 40% voting for Labour. Interestingly, the Independent's readers were more likely to vote Liberal Democrat than those of any other newspaper and the paper did run some enthusiastic editorials in favour of Paddy Ashdown, even if it did finally endorse Labour.
The third cluster of newspapers consists of those with a Labour vote of about 30%, including the Daily Mail and the Times. The Mail was clearly much more pro-Tory than the Times in 1997, although it is interesting that it was much less Conservative than its immediate rival, the Daily Express. Rather more than a fifth of Times readers voted Liberal Democrat - placing it just behind the Guardian in its support for Britain's third party. The most Tory of all newspapers, not surprisingly, was the Telegraph, although about a fifth of the Telegraph's readers voted Labour in 1997.
The most evenly balanced newspaper in relation to the voting behaviour of its readers was the Financial Times. The Financial Times is the polar opposite case to the Guardian in terms of the willingness of its readers to vote, with no less than 39% of them failing to turn out in that election. There is a well-known paradox in political science which states that if people are rational individuals who carefully calculate the costs and benefits of any type of action before they embark on it, then they will not vote. This might explain why the newspaper most closely identified with economic rationality and the market has a readership which is most willing to free-ride on its fellow citizens and not vote.
The final group in the chart are people who do not read a newspaper at all. In terms of Labour voting they were similar to the Sun and Daily Mail readers with roughly the same proportion of them opting for Labour as actually voted for the party in that election.
Overall there is a clear association between voting behaviour and newspaper readership. The top five newspapers in the chart all endorsed Labour and three of the bottom four endorsed the Conservatives. The Times, which sat on the fence, is in the middle. Thus the Financial Times is really the only paper which is out of order and this may be because its readers tend not to vote anyway. There are two possible explanations of these findings: either readers are following the endorsements of their favourite newspapers, or editors are carefully reflecting the views of their readers. If, as seems likely, both things are happening at the same time then clearly editors, or in some cases newspaper proprietors, are exercising real political power.
Paul Whiteley is professor of politics at Sheffield University
paul.whiteley@sheff.ac.uk
