Friday, May 15, 2009

YouGov sub-sample : well I think that answers my question!

I wondered aloud last night whether Labour's collapse in the latest UK-wide YouGov poll really heralded good news for the SNP, or whether the fringe parties would simply benefit at the expense of all the main parties in Scotland. But the detailed breakdown of the poll has now been released, and the Scottish sub-sample for the Westminster vote shows the following -

SNP 38% (+9)
Labour 27% (-1)
Conservatives 15% (-14)
Liberal Democrats 14% (+4)
Others 7% (+4)

As ever there is a huge health warning that needs to be attached to any sub-sample, but it does look potentially significant, especially since YouGov is the pollster that in the last few months has been consistently least likely to show the SNP in the lead for Westminster.

PS. I've just received one of those tedious "hard-working Scottish families love the Labour Party" leaflets - of course that sort of thing has always rung hollow for me, but I suddenly realised looking at it just how out-of-touch it's going to look to the average floating voter in the current circumstances.


  1. James

    My advice - wait for a real all scotland poll. For reason I stated elsewhere, these sub samples of UK wise polls near always have a rural areas sample bias in Scotland that artificially boosts the SNP rating and under-reports Labour's.

    To remind folks, the last stand alone Scottish poll with was on 23 April in the Herald ( With a sample of 1000 and with it 50 plus balanced sample points across scotland) On Westmisnter voting intentions it reported:

    Labour 32% (-5)
    SNP 30% (+3)
    Tory 21%
    Lib Dem 13%
    Other 5%

    This was a good poll for the SNP, with momentum towards it. But the idea that in less than three weeks the SNP is now 11 points ahead of Labour fanciful

    I hope I am wrong. But I'll await for the next all scotland poll - and wait fovrever i reckon for nat bloggers headlines, "Massive slump in SNP vote, as Scots rally to Labour". Because that would be mince, but the logical headline after hyping up this non poll

  2. Aye We Can, I take some of your points (because I've made some of them myself), but for the record, over the last few months I've tried to report every sub-sample - I only missed a couple because I was too busy. I've reported quite a few that have been unfavourable for the SNP - admittedly my headlines have probably been a little more understated in those cases, but I haven't pretended those polls didn't exist.

    By the way, I accept that this sub-sample is not in any way proof that the SNP is 11 ahead points ahead of Labour, but I don't find it at all 'fanciful' that such a surge in support could have taken place in such a short space of time. In these circumstances, three weeks is an absolute eternity in politics.

    I also completely dispute your assertion that sub-samples routinely overstate the SNP's support and understate Labour's - after Glenrothes last year, we were seeing sub-samples putting Labour thirty points ahead. You'd be a brave man to call that an understatement.

    I've no idea whether I'll continue posting about the sub-samples, but if I don't it'll be more down to boredom than anything else.

  3. Jmaes

    Ive no argument with you publishing these figs - they do tell us something, and these figs are certainly encourging, beter than being behind etc. Id just not over egg them for reasons we probably both agree on.

    Ive also just been to the SNP's site and see its majored with a news release on this very poll. This explains - to me anyway - why most nat bloggers are bigging it up, though I accept you have published sub samoples in the past unprompted, and good, bad or indifferent

    Maybe it's just me that has not yet factored in we are in election mode!