“Labour doing better on a low turnout” – surely not?
A new electoral dynamic or an error?
One of the frustrations of yesterday’s public holiday in most of the UK was that the Ipsos-MORI offices were not open and we have yet to see the detail and the analysis of the August political monitor which featured in Sunday’s Observer.
For there was an intriguing line in Gaby Hinsliff’s report that went against all the received opinion on voting behaviour – that Labour do better with higher turnouts. Her comment that the details show that if turnout slumped to 50% then there’d be a “hung parliament” didn’t seem to make sense and have been discussed a fair bit on previous threads.
This was in a poll, of course, that had headline figures of CON 43: LAB 26: LD 17 – the biggest Tory share of any of the firms reporting in August and, with ComRes, the biggest Tory lead. How could such a survey even hint that a hung parliament was possible?
My guess is that there has been a mix-up in the two different sets of voting intention numbers that the firm makes available – the headline ones of those “certain to vote” and the shares of all those in the survey “naming a party”. It is normally MORI’s emphasis on the former that trims back Labour’s position – thus in July Labour’s 16% deficit amongst “the certains” became 8% amongst “all naming a party”.
Maybe, uniquely in this survey for some reason, this wasn’t the case. We’ll have to wait an see.
I’ll update this post when we get the information.
UPDATE 1015
The Ipsos-MORI data is now up and I cannot find anything to support the Observer comments.
I think that she has ascribed the “100% certains” (50% of the sample) to “all those naming a party” (78% of the sample) to reach a conclusion that is the reverse of what the data suggests.