Tuesday, June 11, 2013

ICM poll : Unprecedented surge for the SNP and Plaid?

The Guardian have released the details of the new GB-wide ICM poll on Westminster voting intentions.  Let me say from the outset that I'm a tad sceptical about the astonishingly high combined figure for the SNP and Plaid Cymru, and I do slightly wonder if someone at the Guardian made the error of subtracting the Green vote from the combined vote for 'others', and assuming that what was left could only be the nationalist vote.  The fact that the BNP (which had 4% in the last poll) is seemingly on zero is particularly suspicious.  However, until we hear otherwise, we must assume that the following figures are accurate -

Labour 36% (+2)
Conservatives 29% (+1)
UKIP 12% (-6)
Liberal Democrats 12% (+1)
SNP / Plaid Cymru 8% (+5)
Greens 2% (+2)

I can't check the historical data at the moment because the ICM website is playing up, but a nationalist vote of 8% must be some kind of record.  To put it in perspective, Scotland and Wales have a combined 14% share of the GB population, which would suggest that more than 55% of the electorate in those countries currently support the SNP or Plaid.  OK, that's the kind of freakish result that can occasionally happen when you're relying on very small Scottish and Welsh samples, and shouldn't be taken at face value.  All the same, it would be very unlikely to happen if the nationalists weren't at least doing reasonably well.

On the broader picture, at this stage in mid-term an opposition party heading for victory should have a comfortable double-digit lead.  Labour's slender seven-point advantage does nothing to change my view that the most likely outcome of the next general election is a Tory victory.

UPDATE : The above figures are now confirmed - the SNP are on 7%, comprising a whopping 55% of the Scottish subsample.  Plaid Cymru are on 1% (16% in Wales), while the BNP have indeed slumped to zero.


  1. I would find that difficult to believe. Here is poll for TNS at a combined 5% for SNP/Plaid that appeared yesterday.


  2. I couldn't agree more with Marcia. The poll stinks of random variation; I suspect we'll see a restoration to the usual 9/10pt lead Labour seem to be experiencing next time round. Good to see you've recognised it in your reference to small Scottish/Welsh samples.

    In terms of UK government victories, we all know that FPTP currently skews in favour of Labour; the Tories require a good ~12-13pt margin to secure a workable majority in terms of uniform swing, but Labour require far less. Yes, the majority that could be produced with ~7pts would be slender after the intense 2015 GE run-up, but if you count LD fallout going to Lab/Con and Tory votes going to UKIP, I suspect the numbers would stack up favourably. I therefore disagree in terms of your view to a Tory victory, in spite of the leadership, economic credibility and public trust crises Labour presently faces, which will likely be addressed over the next two years as policy emerges.

  3. Ipsos MORI poll out today:


  4. As I suspected ICM figures should read 3% and not 8%:


  5. Yes, I'm not surprised either - the 'Grauniad' live up to their nickname once again. It seems they can't distinguish between the BNP (on 5%) and the SNP!

    But the Ipsos-Mori figures are much better - the SNP in the mid-thirties in the Scottish subsample.

  6. The tables are now up on the ICM website.

    It seems that 30 voters in the 'North' are planning to vote SNP!!! (See page 6)


    On the other hand, there are also 30 voters in Scotland who want to vote for the SNP. I suspect that somewhere in their analysis, their geography has become confused. Easy to do in the mindset of someone who thinks of Scotland as somewhere remote 'up North'.

  7. OK, we've got ourselves totally confused here. My humblest apologies to the Guardian - they were right all along. The SNP are on 7% (55% in the Scottish subsample), Plaid Cymru are on 1%, and the BNP are indeed on zero. The figures Marcia mentioned are from the May poll (ICM take forever to update their website).

    In response to James Kay, ICM always used to lump Scotland and the North of England together in a "North" region. They still do that, but at least they now supply an additional breakdown for Scotland (that's why the SNP have the same number of voters in "the North" and in Scotland).