Can Labour break through the 32% ceiling?
-
Will new polls confirm whether the turmoil is still causing damage?
When the May Populus survey showing a Labour deficit of 8% came out last week I urged caution because the survey had been taken against a backcloth of heated speculation and talk of a Labour civil war. Often surveys that take place in such a charged atmosphere, I noted, produce extraordinary results.
We should get a better measure of the current state of public opinion over the next eight or nine days when a number of polls are due, particularly the May ICM poll for the Guardian and YouGov in the Daily Telegraph.
If these are still showing Labour at 32% or less then the Blair/Brown’s party will start to look in trouble. Five of the last eight polls have had Labour at this level, one has put the party at 31% and two, including MORI which I attach less importance to, have recorded figures of 30%.
What all the pollsters have been showing is a big rise in support for “others” – particularly the BNP which appears to be damaging Labour more than the Tories.
The significant lead that Cameron’s party has been getting has been based more on Labour doing badly than a big switch to them. The recent average is still one or two points below the peaks reached in the immediate aftermath of David Cameron’s election as Tory leader last December.
The key numbers in all polls are the voting intention ones for it’s not always appreciated that these are often calculated on a different basis than other poll findings such as how people rate Blair. ICM and Populus scale down support based on how people respond to the “likelihood to vote” question. Both have applied the same turnout-linked adjustments with Brown-Cameron voting intentions.
YouGov normally does not apply this corrective measure – except in pre-election polls. Generally the turnout question favours the Tories, cuts back Labour and leaves the Lib Dems where they were.
I tend not to attach too much importance to MORI because it does not weight by past vote recall and is more vulnerable to sample error. I’m also cautious about the new Mail of Sunday’s pollster, BPIX, which fails to respond to my repeated requests for further information on how its figures are arrived at. For its last poll it did supply information to Anthony Wells of UK Polling Report but not me
The General Election betting markets show that punters are still rating Labour ahead of the Tories – but only just. Maybe that will change with new polling evidence.
Mike Smithson