YouGov’s monthly(ish) Welsh poll for ITV Wales is out. It was conducted after the results of the Welsh referendum were known. Topline figures with changes from last month are:

Constituency: CON 20%(-1), LAB 48%(+3), LDEM 7%(nc), Plaid 19%(-2)
Regional: CON 20%(nc), LAB 45%(+4), LDEM 5%(-3), Plaid 18%(-3), UKIP 5%(+1), Green 4%(+2).

I’ll add all my normal caveats about being cautious about a single poll showing movement, but there appears to have been a slight shift towards Labour since the referendum. On a uniform swing, my projection is that it would be enough to give Labour an overall majority in the Welsh Assembly, producing 33 seats for Labour, 14 for the Conservatives, 10 for Plaid and 3 for the Liberal Democrats.

Also notice the Liberal Democrats way down on just five in the regional vote, compared to 12% at the last Welsh Assembly election. On that level of support they probably wouldn’t get any regional seats at all, they avoid wipe-out because of some safe constituency seats. UKIP will probably be happy to see themselves equal to the Lib Dems in the regional vote, though it wouldn’t give them any seats (though not far off – if they get to 6% then on a uniform swing here’s a chance UKIP could get a regional seat in South Wales East)

71 Responses to “YouGov Wales poll would give Labour a majority”

1 2
  1. Prove that Old Nat is okay…

    or I spill the beans!

  2. Hmm….I think the polls are showing a LD massacre in May. However, I don’t think they will cut and run to a Confidence & Supply……..a grim looking Nick will (with added gravelliness in his voice) come out on the friday and encourage stoicism amongst his defeated ex-cllrs and their activists “We have lost a battle, but not the war for Britain’s economic recovery….hard decisions…when the going gets rough in Govt the LDs are a tough bunch and can take it….yes, we can take it!”
    All Dunkirk spirit stuff. That will be the LD narrative, and any dissenters will be portrayed by the Cowley Street machine as “distraction whingers”.
    For this reason an end to the coalition is an extremely unlikely scenario.
    @ Eoin
    David Laws and extra roving Cabinet post? Perhaps, but maybe Cable for the chop – he’s lost his gloss, and much following in the party with his gaffs – could he be safely sacked to make way for Laws as Business Secretary?

  3. @Tony Dean – “… Cowley Street”

    Not for long, they are on the look out for a new HQ since the party went into the red.

  4. @ Billy Bob

    4 Cowley St. to go? – Oh, what a shame – I rather liked the old building when I was employed there back in the 80s!

  5. Tony,

    DC has [reportedly] agreed to a 6th LD position at cabinet. Technically Laws would not have any portfolio. So Cable to say.. Old rule applies.. you keep your friends close but…

  6. Tony Dean

    “will be portrayed by the Cowley Street machine.”

    Except that they’ve run out of cash with no Short Money and have to move out.

    Perhaps the “Railway Cuttings, East Cheam” machine?

  7. @Tony Dean – “… 4 Cowley Street”

    It is already on the market – “new instruction”.

    Was the choice for HQ down to Roy Jenkins I wonder?

  8. Has any one seen the reports that the Saudi’s have been shooting protesters

    We are living in interesting times

    All those traders who have taken options on 200 dollar oil will be sleeping soundly tonight

  9. Now now…. all those champions of interventionism…. who wants a No Fly Zone over Saudi Arabia? ………………? Didn’t think so.

  10. Anthony,

    Looking at the PDF, I see there’s another very large weighting up of Lab party identifiers (289 to 360) as with the recent Scottish poll, but the PDF doesn’t show party id or newspaper weightings for Wales.

    Are they available separately, perhaps? If so, would you have a link?

  11. @Eoin,

    If Saudi reached the point Libya has, then it’d have to be considered. I don’t think Saudi will though. They may be as greedy as Gadaffi but I don’t think they are in the same ball park on the bonkersness.

  12. Zawiyah

    h ttp://

  13. 11 point lead. Is this an outlier?

  14. No response from anyone. Are you all in shock? Is this the result of the LibDem conference and their 2.5m steel and concrete fence?

  15. @ Eoin

    “David Laws [once cleared] is set to return to the heart of government.

    De jure, LDs are only entitled to 5 Cab posts and it will remain that way, but a sixth post will be created de facto as Laws will have the right to attend cabinet. His job will be to promote a commonality of purpose [at the cab office.]”

    If you’re right about this, I’m glad. I’m a fan of David Laws (even though I have to say that I think many of his economic ideas are kinda half baked). I think the fact that he had to leave the government was silly. Yes, he technically broke the rules but he did so in a way that was adverse to him as opposed to beneficial.

  16. @ Liz, “are you all in shock”

    No – you are on the wrong thread!

    Final thought on the LD travails before I hop over and have fun on the 11point lead!

    If Cable is sacked to make way for Laws you would have an easy rallying point for dissent. I could even see Cable defecting (he used to be my MP and he’s a really decent guy). You can sack people for gaffes but actually his gaffes made him more of a hero in the eyes of his party.

  17. Pretty fascinating for May.

    Aberconwy will be very interesting – should be a straight Plaid v Con fight, but if these figures are right, Lab might come into play.

    Montgomeryshire – Cons mights pick this up? Bit of a suprise when old Opik went last May, but Lib vote looks shaky.

    Cardiff Central – I think the Lib vote may well evaporate here, Lab gain? Libs will have done extremely well if they hold this one, especially with incumbent gone, pissed off students etc.

    Clwyd West – Lab challenging Con strongly here?

    All very interesting…..

  18. Anthony,

    You said “I wouldn’t pin any hopes on regional cross-breaks. They have small sample sizes and aren’t internally weighted”


    Is there likely to be enough difference between the individual respondents in that regional element of the sample to mean that we can get added confidence by looking at the regional cross-breaks over the last few polls?

    This wasn’t the first poll where the regional cross-break showed the Greens doing well in South Wales Central. Are we just getting the same individuals, who may -or may not- be typical of voters in the region, or are we getting evidence that might be cumulative?

    Meanwhile, as a GP insider, I doubt the Green Party are “targetting their resources” (to quote James) sufficiently to have caused the extent of the difference in polling that this poll says: the size of the region is just too big. My guess is that a significant part of this is people who voted LD at the General Election turning against them (note that South Wales Central includes all of Cardiff Univ, Glamorgan Univ, University of Wales Institute, Royal Welsh College of Music and Drama and the University of Wales College of Medicine). And where do people upset with both the coalition in Wales and the UK coalition turn?

  19. Ben –

    Aggregating up the data for multiple polls gets over the problem of small sample sizes (so it is better than looking at single polls!), but doesn’t get over the problem of no internal weighting.

    Samples are designed and weighted to be representative of Wales as a whole, not necessarily within each cross break. Of course, with luck it should all work out in the wash, but there’s no guarantee of it.

    It’s possible for example, for a sample to have the correct proportion of Labour identifiers overall, but to have too many in one region and too few in another. It pays, therefore, to be wary of cross breaks, especially in something like voting intention where differences of just a few percent can matter very much.

  20. Thanks Anthony.

    I see what you are saying about not weighting within regions.

    Your post brought a smile to my face: yesterday morning I was giving some of my students a practical demonstration of how to apply demographic weighting to survey responses (spreadsheet open, typing formulae in…).

1 2