As well as the regular GB poll, YouGov have released new Welsh and London polls this week.

The London polling for the Evening Standard is here and has topline figures of CON 32%, LAB 42%, LDEM 7%, UKIP 10%, GRN 8%. Despite its Conservative mayor London tends to be more Labour than the country as a whole – at the last election the Labour party were two points ahead of the Tories in London, compared to the seven point Tory lead across Great Britain. This means a ten point Labour lead in London is a four point swing from Con to Labour, the equivalent of a one point Labour lead in a GB poll. In other words, the swing to Labour in London is pretty much the same as in Britain as a whole.

The Welsh polling for ITV Wales and Cardiff University is here and has topline figures of CON 23%, LAB 37%, LDEM 6%, PC 10%, UKIP 16%, GRN 8%. Compared to the general election result this is Labour up 1, the Conservatives down 3 – a swing of 2 points (so actually a smaller swing to Labour than in Britain as a whole). Roger Scully of Cardiff University’s analysis of the poll is here.

419 Responses to “YouGov London and Welsh polling”

1 2 3 4 9
  1. I mean the change in London’s VI seems to be about average for the UK.


    It looks like they are and I still think UKIP will fall back a bit benefiting the Tories so in a few weeks or a months time we could be looking at a 35%+ VI for the Tories.

    Ok I’m done for tonight. Up early and no not for a trip down to the Mendips to see a Bishop.

  3. @Neil A – understood.

    As with all this swing thing, it’s all about timing.

    Syncopation. Start at the wrong time, and nothing swings.

  4. @Oldnat,

    I don’t think the high birth rates are a new thing!

  5. Earthquake in Winchester.

    Link in the paper that cannot be linked to.

  6. Ukip and LD falling, as all we see in the media is Lab and Tory new policy launches. Polls will change a lot in the next 10 weeks, and maybe settle after that, once the debates are over.

  7. @Alec

    But no one believes that the Greens will be in a position to implement any of their madder policies post May. The Green voter is probably the same type of voter that used to vote Lib Dem – quite idealistic so possibly their self-image is more comfortable being a Green than any other party.

  8. @allan Christie

    Agree, although my bet is months rather than weeks … Tory average for January will likely still be below 33%, and no reason to think there will be anything other than a very slow drift if anything until April when, well anything could happen to the polls quite frankly

  9. @DW living in Cardiff West I’d not quibble for a moment with your description of your own constituency but I do doubt your assessment of the 25%; people in Powys born in English hospitals is a factor for that county but not for the high incomer/immigrant levels in north Wales – where there are large maternity units in Bangor and Wrexham or places like Ceredigion: we only have very limited knowledge of how country of birth affects UKIP support but that question has been asked about support for devolution and the correlation between country of birth and support (and opposition) to devolution is well established.

  10. Colin

    That could possibly be at least partly down to UKIP being a lot less forthcoming about policy

  11. Chris in Cardiff @DW

    “Country of birth” is, however, simply an objectively measurable proxy for national/cultural/community identity.

    By itself it is irrelevant – witness the SNP MSPs born in England.

  12. @ OldNat agree on all counts. Witness Lloyd George born in Manchester and Dafydd Wigley born in Derby; I do think it’s relevant for Wales, albeit imperfect

  13. I’ve been looking for some stats around how ethnicity will change between 2011 and 2015, and have finally found some for London here

    The 2014 set of those stats have
    – White British -3%
    – Black other +1%
    – Other Asian +1%
    – Other +1%

    Looking at the weights in the Yougov London poll, those look like the 2011 census weights.

    I’m sure that change adds 1% to Labour VI if the correct 2015 weights are used.


    “That could possibly be at least partly down to UKIP being a lot less forthcoming about policy”

    Would that include or exclude “100 reasons to vote UKIP” on their website?

  15. The other thing those stats give is a breakdown of the BME population by age range.

    For the 2015 projection
    18 – 54%
    19 – 54%
    20 – 49%
    21 – 46%
    22 – 43%
    23 – 42%
    24 – 42%

    So on average you can say around 45% of that 18-24 age bracket is BAME.

    BUT – looking at the 18-24 crossbreak, 23% are voting Green – we know Green voters are younger!

    But are BAME voters really voting Green, or is this just lots of white british young voters? The BES data doesn’t reflect any Green move amongst BAME voters.

    So I think this London poll could be wrong, it is understating Labour, and overstating Greens, it needs more BAME voters in that young age group to be accurate. Having to upweight both young and BAME voters is leading to flawed results.

  16. @Old Nat
    “Given the massive upweighting of minority ethnic (hate that word!) voters, I wonder how accurate that London poll is – even if it is more likely to be better than one that didn’t bother weighting for that.”

    If only “than one” were just one.

    It’s reasonable to assume that ethnic minorities are massively underrepresented as YouGov panel respondents in general, and not just in London. And no attempt is made in national YouGov polls to correct for that.

    And who’s to say that most other pollsters, who similarly don’t weight for ethnicity, have any better raw samples? If YouGov are struggling surely others will be too.

    So there could be a significant cohort of the population – ethnic minorities – whose opinions are for the most part being ignored in most national polling.

    I think Mori might be an exception on weighting – I recall seeing a crossbreak for ethnicity in their issues polling.

  17. Phil Haines

    IIRC Anthony said that their research indicated that ethnicity wasn’t much of a factor outside London.

    While weighting for every possible factor might seem wise, presumably not if it doesn’t produce change.

  18. Stutter

    Includes. The Greens make all of their policies easily available online.

    Note how they need a whole website for their policy rather than a single blog post.

  19. It’s interesting to look at the 12 YouGov polls, taken over the last fortnight or so. The Labour vote has hardly shifted at all. Take a look (oldest first, latest last): –


    If anything, after a sequence of 32,32 and 30, it’s picked up slightly and solidified in the 33-34 range. Now let’s look at the Tory vote over the same period: –

    32,32,32, 31,31,32,32,33,31,32,34,34.

    Flatlining, then two sudden slight upticks. No obvious net gain from Labour, so where has the increase come from?. UKIP returnees? Let’s see how UKIP have polled over the same period:-


    The very faintest sign of a downward trend maybe, but only slight and they were scoring 14s two weeks ago when the Tories were on 32. There doesn’t appear to be any great correlation between the Tory and UKIP VI, although it needs to be kept an eye on as time goes by.

    So, if the two consecutive Tory 34s aren’t as a result of a downturn in either the Labour and UKIP VI, where has their increased support come from. DKs coming off the fence, perhaps? Difficult to tell amongst the churn, but there’s no telling evidence of Labour voters switching or UKIP voters returning.

    Could be a bit of MOE and, if things really are all square, then maybe a run of narrow Tory leads is balancing out the run of narrow Labour leads that came before.

  20. Crossbat11,

    I think that the changes are too small to be interpreted. What one has is a situation where the parties are roughly even in VI; any more than that is really stretching the data.

  21. If I recall correctly, on a five-poll rolling average, this second Tory one-pointer replaces the Tory two-pointer, meaning the Labour position is a slight improvement on yesterday on that measure.

    The conclusion is that they are neck and neck on Yougov.

  22. It’s the closest we’ve come to the dreaded “C” word for a very long time…. Another false dawn?

  23. @Old Nat
    Yes I’m aware of AW’s position. It may though be one borne out of necessity not because ethnic minorities are an insignificant share of the GB population, but because they are such a small share of the YouGov panel.

  24. @Bill Patrick.

    Individually polls may have a margin of error of maybe +/-3% but it is a common misconception in the press that one can use a similar margin of error when assessing the cumulative findings of multiple polls.

    Looking at the shift in VI across the 6 polls published in the last 24 hours it is very likely something has changed although the specific shares of the vote are uncertain.

    The question is why and whether it is a blip.

  25. Today’s YG Scottish crossbreak

    SNP 44% : Lab 28% : Con 23% : LD 3% : UKIP 0% : Grn 3%

    Mean of last 18 YG Scottish crossbreaks

    SNP 42% : Lab 27% : Con 17% : LD 5% : UKIP 5% : Grn 4%

    Looks like the “shy kipper” has been filleted. :-)

  26. Here is the YG data for Monday and Tuesday by 2010 party ID:

    On;y two days data though!

  27. @CMJ
    Superficially it looks like UKIP drifting home.

  28. 4-1 to the Tories over the past 2 days, a [] trend?

  29. @Adam Kennedy

    Not if you’re a Tory!

  30. Hmmm–one is tempted by this mini -crossover to start thinking of the “S” word.

    Early days with, as Bill says, not enough distance from a MOE effect as yet.

    Still-not a bad poll:-)

  31. @Oldnat

    LiS edging towards 30% but no sign of slippage in the SNP vote.

    With just 100 days to go…

  32. @crossbat11

    Very interesting trend analysis of recent YouGov polls – the lack of a clear source for the Tory uplift reinforces that this might just be normal random variation rather than a move out of the 31-33 band, but I think the Tories will probably drift up in VI as the election approaches and people increasingly compare a pretty average leader in Cameron to a pretty awful leader in Milliband

  33. Yesterday’s YouGov weighted data was Con 34.3 Lab 32.5. Today the figures seem to be Con 33.5 Lab 32.8

  34. @John Politico

    We keep hearing how Tory VI is due an uplift at this stage in the electoral cycle (largely based on swingback theory). The problem I see for the Tories is that at no time over 4+ years have they been able to scale 35%+ for any longer than a couple of days, whereas Labour now appear to have stabilised at 32-33%.

    The conventional wisdom is that if Labour polls, say, 33%, the Tories will need 35%+ to end up the largest party. Of course, the Tories could continue to govern as a minority government with LD+Ukip+DUP support on less than this percentage but they would not be the largest party and Lab+SNP may be able to block any Tory legislation.

    So the Tories probably need 35%+.

  35. @RAF
    I still think that the shy Tory phenomenon and a bit of differential turnout, will lift the Consevative vote by 1% to2% when it comes to the actual vote. This could mean that they are already on 35%+. All just imo of course and based on just a few polls.

  36. the momentum, even at this glacial pace, is with the conservatives…labour will be hoping the election comes quickly…the tories need all the time they can get.

    labour’s performance has been pretty abject in the last few months. not really having a narrative on the economy is not great for them.

    Bill Patrick was absolutely right to say that miliband had either a) to be Blairite centrist or b) be an out and out left winger and fire up his base…he has done neither, with the consequence that labour will do well to get 33% of the vote this May.

  37. RAF
    The problem I see for the Tories is that at no time over 4+ years have they been able to scale 35%+ for any longer than a couple of days, whereas Labour now appear to have stabilised at 32-33%.

    not quite true. pre omnishambles budget in march 2012, so less than 3 years ago, the tories regularly got 35%+…

    not sure labour has stabilised, they are still on a downward trend.

    I think in terms of seats more than VI, so looking at the seats, the way Iain Dale did, I can’t see how the tories get more than 285, which even with the liberals on about 35 would not be enough to control the house.

  38. crossbat11 last night
    “It’s interesting to look at the 12 YouGov polls, taken over the last fortnight or so. The Labour vote has hardly shifted at all. Take a look (oldest first, latest last): –
    If anything, after a sequence of 32,32 and 30, it’s picked up slightly and solidified in the 33-34 range. Now let’s look at the Tory vote over the same period: –
    32,32,32, 31,31,32,32,33,31,32,34,34. Flatlining, then two sudden slight upticks.”

    The same 12 YouGov polls, taken over the last fortnight or so, put in order of size, looks like random error to me. If anything, both a bit too steady.
    30,32,32,32,33,33,33,33,33,33,34,34 average 32.75
    31,31,31,32,32,32,32,32,32,33,34,34 average 32.17

  39. GRAHAM

    You do realise what the MOE for these polls is-don’t you?

    Decimal points are meaningless, spurious” accuracy “

  40. ‘Politicians should leave it to the professionals.’

    And in education!

  41. Interesting article in the Times about Tsipras & his new finance chief.

    The latter is a graduate of Essex Uni-he is charged with the negotiations with the Troika. Struck by his determination to stay in the EZ. That seems like a weakness in the cards he holds to me.

    Tsipras made some pointedly anti-German gestures after his appointment. The Times says he is pro Russia in the Ukraine issue & wants sanctions lifted.

    He has said that his new Tourism Minister will seek to reduce the attractiveness of All-In Package holidays because they harm local businesses. Tourism Industry reaction is predictable !

    Interesting times .

  42. LURKER I don’t know who you refer to.

    I referred to Simon Stevens-CEO of NHS England.

  43. COLIN

    “And as for the two Labour vets who cling…”

    It is no good being obtuse when you can’t edit your previous posts.

  44. With current polling now consistently showing a small Tory lead and the experts themselves saying that the polls are “too Laboury* I would probably put the true Tory lead at 3%.

    EM and Labour are looking like also-rans.

    I also think the Labour mantra….

    There were five in the bed and the little one said –

    “These NHS cuts are getting a bit much” has probably ran its course now.

  45. LURKER

    I wasn’t being obtuse-just not following your train of thought properly.

    I see what you refer to now-cast into the uttermost darkness then -eh?

  46. @allan Christie, @crossbat11, @RAF

    I think swing back theory supports a gradual drift back to the Tories over the coming months, but I think it is the Cameron vs. Milliband effect that makes me believe it is more likely than not — I agree Tories will need well over 35% to have a chance of forming a government

    What is impossible to know at this stage is whether Farage and UKIP will start to fade away or actually thrive on the publicity they will have and the benefit of a simple, populist message

    The problem for the Tories is if the polls do look like they are moving in favour of a Tory win, the “go to bed with Nigel and wake up with Ed” message won’t resonate … which is why I think the Tories just cannot get over 38% even if Labour meltdown

    As it stands, even with some drift back to 36/37% it is 50/50 on whether Ed or Dave will be the next PM for me

  47. RAF

    “LiS edging towards 30%”

    Not really. Daily figures in Jan are
    30, 28, 28, 24, 25, 28, 28, 28, 22, 24, 31, 33, 26, 31, 23, 25, 32, 28,

    and the rolling mean score has gone from 29 at the start of Jan to 27.4 now.

    As you would expect with samples of c.150, the numbers bounce around a bit, but there has been little change for anyone throughout January.

  48. @Mr Nameless

    Did you see Eamonn Burgess when you were in Norwich?

  49. Latest odds/spreads on GE:

    Most GE Seats Betting Odds:
    Tories 5/6
    Labour: 10/11

    Spread Betting on GE Seats Won:
    Tories: 279-285
    Labour: 278-284
    LD: 27.5-29.5
    UKIP: 8.5-10.5
    SNP: 32-34

  50. If we are looking at ethnicity we should really ask how Scots in the rUK are voting. Just to keep it interesting!

1 2 3 4 9