Was it their “17% Labour margin” that made the SNP prices attractive?
Yet again an election has shown the pollsters up and reinforced my default position when assessing all voting intention surveys – “always assume that Labour is being overstated“.
The whole atmosphere in the lead up to yesterday’s Glasgow East by election was affected by the Scottish Progressive Opinion survey that had Labour 17% ahead. I was highly sceptical about their methods right from the start and it was an examination of their detailed data that led me to conclude that this was a 50-50 by-election.
Yet even though the polling industry as a whole has an appalling record of Labour over-statement in by-election surveys few pundits and punters and pundits had the confidence to dismiss PSO’s 17% Labour margin out of hand.
Thus we saw the Politics Home PHI 100 panel give it to Labour by a staggeringly awful 86%. They really ought to read PB!
The markets saw Labour tighten to as much as 2/5 after the PSO survey came out and the party remained heavy odds-on favourite until the reality of the count started to emerge at just after midnight.
It was this poll, I believe, that drove media opinion and led to all those stories that Labour were going to squeeze home.
Finally I should say well done to the BBC. The coverage overnight was really excellent – maybe our messages are getting through.