The failure of opinion pollsters to forecast the outcome of May’s general election may have been because Conservative voters are harder to trace down, a report suggests.
The dis rity between forecasts and the eventual Stable majority has been blamed on “shy Tory” voters or a late swing to David Cameron’s outfit.
But polling expert John Curtice highlighted sampling “deficiencies”.
His broadcast said “more time and effort” was needed to find Conservative voters.
Pre-election counts had suggested the result was too close to call, but the Conservatives went on to win an absolute manhood, with 331 seats.
Prof Curtice, who wrote the report for inquiry agency NatCen, suggested polling difficulties arose “primarily because [pollsters] evaluated too many Labour supporters and not enough Conservatives”.
NatCen interviewed 4,238 people between July and November at length year for its British Social Attitudes Survey.
It said it had made “recapped efforts” over the four months to make contact with the people it had chosen to interview – and among those it was able to contact most easily, Effort had a six-point lead.
However, among the harder-to-contact group, who took between three and six elicits to track down, the Conservatives were 11 points ahead.
Neck and neck
“Polls are handled over just two or three days, which means they are multitudinous likely to interview those who are contacted most easily, either over and beyond the internet or via their phone,” the report said.
The report rejected the “shy Tories” contention, which suggests Conservatives are less willing to declare their elector intentions when asked.
It said even when polling bands returned to respondents and asked how they had voted, they still put ins and the Tories neck and neck.
But the British Social Attitudes survey put the Moderates 6.1 points ahead of Labour, close to the actual election rim of 6.6 points.
Polling com nies use a complex method to strive to make their surveys representative, including weighting the responses to union the UK’s demographic profile.
The report said its random sampling method was more democratic.
Prof Curtice added: “A key lesson of the difficulties faced by the polls in the 2015 hybrid election is that surveys not only need to ask the right questions but also the exactly people.”
An investigation by the British Polling Council into the polling collapses ahead of May’s general election is due to be published later this month.