Red-faced pollsters get it wrong again
Only two polling analyses accurately predicted a hung parliament - because companies wrongly adjusted their figures as a result of being "burnt" by past failures, experts say.
An analysis by YouGov last week pointed firmly towards a hung Parliament but was widely dismissed as inaccurate. A separate poll by Survation put the Conservatives one point ahead of Labour - just two short of the actual result.
In 2015 pollsters across the board were left red-faced after failing to predict David Cameron's win over Ed Miliband. Major firms admitted they had relied too heavily on a relative over-representation of politically engaged young voters, many of whom did not vote.
This time, polling companies over-compensated in their adjustments, YouGov's former president, Peter Kellner, said. In many cases the raw figures "pointed to a bad night for Theresa May".
"Had they stuck with these numbers, they would have redeemed their reputation," Mr Kellner wrote.
"Instead, burnt by their failure to predict David Cameron's victory two years ago, they adjusted their data in the belief that many young Labour supporters would end up staying at home."
Ben Page, Ipsos Mori's chief executive, said: "We adjusted downwards the turnout of the young using the British Election Study, which goes back to actual electoral rolls in town halls and checks that the people in its large sample did or did not actually vote.
"What that showed in 2005 and 2010 is that the young do disproportionately overclaim [about their intentions to vote]. We applied that thinking we would make our polls better and of course it had a net effect of understating the Labour vote this time because this time they did vote."