Reappraising the polling that got rubbished by the so called experts
Last night I watched again the Question Time leaders’ special with Corbyn and TMay from the University of York that took place on June 1st exactly a week before the election. On the night itself the PMs performance was well received particularly by right wing commentators.
The only polling on this event, carried by Survation on the Saturday, had TMay doing badly and Corbyn doing well. The survey , it’ll be recalled, got roundly dismissed because it committed the great sin of not fitting the prevailing “it’s going to be a Tory landslide” narrative.
The high proportion of the sample who said they’d watched was compared with the official audience figure was seized on by the Survation sceptics and used it to dismiss the findings about the leader performances and the huge narrowing to 1% of voting gap that it found.
What wasn’t appreciated was that 72% of those sampled had said they’d had seen or heard something about the debate. This wasn’t an audience figure.
What was really damaging to TMay was the widespread discussion and distribution on social media and news clips of the “no magic money tree” response to the NHS nurse who talked about her effective pay cut. Thus the actual programme “reach” was substantially larger than the official figures.
We now know, of course, that Survation is the only pollster that came out of GE2017 with its reputation enhanced and that its findings on the major set piece of the campaign should be taken seriously. This is from the Survation report of its survey for the Mail on Sunday:
“… In total, 72% of respondents had seen or heard something about the debate.
Among this group, 36% said the event made them more likely to vote Labour, vs 24% who said it made them less likely to vote Labour, a difference of +12.
Among the same group, 32% said the event had made them less likely to vote Conservative, vs 24% who said it made them more likely to vote Conservative, a difference of -8… ”
Other post QT special polls found a huge uplift in young voters saying they were certain to vote some of which can be put down to the programme and how it was reported and discussed. This is a trend chart from ICM’s campaign polls the certain to vote turnout responses from the 18-24 segment.
The young overwhelmingly went for LAB over CON. Look at the turnout gap that opened up in those final two post QT polls. This move didn’t impact on ICM’s voting numbers because, of course, the firm was cutting back young respondents to their GE2015 level.