Sample method deficiencies might be to blame for election result poll failure

The failure by opinion pollsters to accurately forecast David Cameron's election win may not be due to "shy Tories" or a late swing to his party but because their sample methods may have resulted in too many Labour supporters being surveyed, a report has indicated.

Polling expert professor John Curtice suggested that studies carried out in the run-up to the May 2015 election, which put Labour and the Tories neck-and-neck, had "deficiencies" in their sampling which weighting failed to correct.

Prof Curtice's report suggested that Labour voters were more easy to contact than their Tory counterparts, giving the party a six-point lead over the Conservatives among those who responded to a survey at the first time of asking.

Among those who were only interviewed after between three and six calls were made, the Tories enjoyed an 11 point lead.

"In short, those who were most easily interviewed by British Social Attitude (BSA) interviewers appear to have been more likely to support Labour and less likely to support the Conservatives to a degree that cannot be accounted for the by the social profile of these respondents," the report said.

"If indeed this group in any way mimics the kind of person who was most likely to respond to the polls, then we can begin to understand why the polls might have overestimated Labour's strength."

The random sampling method used by the NatCen Social Research BSA produced a result that closely replicated the actual outcome of the election.

Asked how they voted, the 4,238 respondents to the BSA put the Conservatives 6.1 points ahead of Labour, close to the actual result of a 6.6 point lead for Mr Cameron's party.

The BSA's respondents are chosen by random sampling, and the study made repeated efforts over the course of four months to make contact with those selected for interview while opinion polls are carried out far more quickly in order to provide a snapshot of the political picture.

Prof Curtice said: "Opinion polls are intended to provide their journalistic clients with a relatively inexpensive way of securing a reading of the very latest political weather.

"They need to be inexpensive because of limited budgets, while a timely reading is often wanted because of a wish to establish whether a recent political event or development has changed public opinion - even though few such events or developments prove to have any electoral consequence.

"As a result polls are conducted in a way that does not fully meet the requirements of random sampling. A key lesson of the 2015 election is that, as a result, they run the risk of failing to take the political temperature correctly.

"Their approach at that election resulted in deficiencies in their samples that subsequent weighting and filtering of the data failed to correct.

"Whereas the polls still largely put Labour and the Conservatives neck and neck even when they asked people after the election how they had voted, two major post-election surveys that used random sampling, BSA and the British Election Study, have both been able to replicate the Conservatives' 6.6 point lead reasonably accurately."

Prof Curtice, senior research fellow at NatCen, added: "A key lesson of the difficulties faced by the polls in the 2015 general election is that surveys not only need to ask the right questions but also the right people. The polls evidently came up short in that respect in 2015.

"BSA's relative success in replicating the election result has underlined how random sampling, time-consuming and expensive though it may be, is more likely to produce a sample of people who are representative of Britain as a whole.

"Using that approach is crucial for any survey, such as BSA, that aims to provide an accurate picture of what the public thinks about the key social and political issues facing Britain and thus ensure we have a proper understanding of the climate of public opinion."

:: The 2015 British Social Attitudes survey consisted of 4,238 interviews carried out between July 4 and November 2, 2015.