It is true that a large proportion of the electorate refuses to take part in opinion polls. In face-to-face surveys, this is hard to quantify, because people cross the street, or do not answer the door.
In telephone surveys, the refusal rate is about 40 per cent. Of those who do agree to take part, up to a further 40 per cent initially refuse to say how they will vote, replying "don't know", "won't vote" or "won't say". But, when pressed ("Which party are you most inclined to vote for?"), this falls to 23 per cent.
Unlike the last election, however, none of the pollsters except MORI take "don't knows/won't says" at face value. Harris, for The Independent, finds that most of them will still say how they voted at the last general election, and assumes that those who say they voted Conservative last time will do so again this time. In last Friday's poll, this increased the Tory share of the vote by three points and reduced Labour's by two - and cut the refuseniks to 16 per cent.
This is a rough-and-ready adjustment, because some of the don't knows will be genuine. But it also leaves out some reluctant Tories - those who were too young to vote last time, or who say they did not vote or they cannot remember.
What then of the 40 per cent who will not take part in the first place? Surely there is an army of hidden Tories lurking there? Not according to Nick Sparrow, managing director of ICM, the polling company which does the Tory party's own surveys and consistently produces the lowest Labour figures.
"What you've got to worry about is not high or low levels of refusal, but the reasons for it - 'I've got the tea on", or 'I'm looking after the kids'," he says.
He is satisfied that the adjustments he makes to the samples he gets take into account the "shy Tory" factor.
But then, if the Tories took polling evidence seriously, they would have given up long ago.Reuse content