British general election (17): how did all the polls get it so wrong?
The failure of all the polls to predict the actual outcome of the General Election has rightly caused many questions to be asked. There is to be an inquiry commissioned by the British Polling Council into what went wrong.
But this is a wider problem than Britain, In other countries, like the USA and Israel, polls have recently got things badly wrong and some hard thinking is taking place.
May 10th, 2015 at 3:55 pm
I am not sure the polls were so wrong, but I suggest the following:-
(a) Ukip by night – some might-have-been Labour voters wavered over to Ukip in the polling booth. It may not have altered the result much but did account for Ukip’s 4m vote share.
(b) differential turn-out : pollsters cannot truly guage whether a promise will translate into a vote. I have always though the core Tory vote, such as it is, is more resolute in actually voting than the core Labour vote
(c) it is possible that though Labour’s postal vote machinery is much better than it was a decade or two ago, the Tories have quietly got theirs slicker too. I look forward to receiving statistics on this locally.
Polling only needs a small error margin to produce very odd results. Scotland was correctly forecast as a near clean sweep. Possibly pollsters didn’t look closely enough at Labour and Tory marginal seats in England.