150429 Election campaign polling
  • From 2010… The great battle for hearts and minds has entered its final week – but are hearts actually captured and minds actually changed? The above graph shows the Tory lead over Labour in YouGov’s daily opinion polls during the last election campaign. At first, this pink line seems to oscillate wildly, from a high of ten points to a low of minus three. And yet, the longer you look at it, the flatter it appears. Even with the great adjustment caused by the first television debate Cleggmania, the dotted trend line for the entire campaign is only slightly tilted. The gap between the two parties started at eight points and finished at seven.
  • …to now… So far, this current election campaign has been even flatter – both literally and in the opinion polls. The Tory lead over Labour has ranged from a single point to minus three, which is practically all within the margin of error. The resultant trend line is again almost horizontal. Could the campaign end as it began, with a Labour lead of two points?
  • …it’s all the same story. All of which chimes with the idea that election campaigns – all the spending, all the handshakes – don’t really make much difference to the final result. One of the few recent elections that really bucked expectations was the Conservatives’ victory in 1992; but, as I explained in another To The Point post, that was less to do with the campaign itself, and more because those expectations were wrong in the first place. As Lewis Baston put it in an excellent article for this site, “Most election campaigns produce results that are more or less what a reader of opinion polls and other electoral data might have anticipated a few months in advance.”
  • And yet 1) The expectation at the start of this campaign was for a very close election result, and the campaign surely won’t change that. But, of course, a close election magnifies even the smallest of changes. The polls may shift by just fractions, but fractions could mean the difference between a Conservative-led or a Labour-led government after 7th May. And that’s before we consider the changes that are going on beneath the headline figures. As Peter Kellner pointed out in a recent commentary piece, “In the past fortnight, around three million voters have switched from one party to another, with a further three million moving to or from the ranks of the don’t knows.” Something has happened to alter their intentions.
  • And yet 2) Besides, election campaigns can make a difference beyond the result itself. Stuff happens in those frantic few weeks – Nick Clegg’s tuition fee pledge, Gordon Brown’s off-camera catastrophes – that, thanks to the media’s heightened coverage and the public’s heightened interest, lingers longer than it would do otherwise. These effects can be difficult to quantify or even to qualify, but they’re there. Numbers can never tell us everything.