YouGov London poll

I’m a little late with this, but just to flag up that yesterday’s Times had a new YouGov poll of London. Topline voting intentions there are CON 34%, LAB 42%, LDEM 8%, UKIP 9%, GRN 6% (tabs here).

I sometimes hear a sort of London as the weathervane of the country argument, as London goes so the country goes. It doesn’t, despite a Tory mayor it’s Labour leaning compared to the country as whole, and it doesn’t necessarily swing in the same way as the rest of the country either. In 2010 while the Conservatives were seven points ahead in the country as a whole, Labour won in London by 2 points, and the Conservatives managed a smaller swing in London than elsewhere.

This poll suggests a 3 point swing from Con to Lab in London, the equivalent of a 1 point Conservative lead in national polling, so actually a little less than national polls are currently showing. It could be the Tories are doing a little better in London, or it could just be ordinary sample error – as ever, it’s just one poll and shouldn’t be overinterpreted. A three point swing in London would net Labour three or four gains from the Tories – Hendon, Brentford and Isleworth, Enfield North and, right on a knife edge, Croydon Central.


396 Responses to “YouGov London poll”

1 5 6 7 8
  1. Breaking news: Ed Miliband has announced the official Labour Party position is that That Dress is white and gold, not blue and black.

    Could this be a game changer?

    @ Anthony,

    I think you know what YouGov’s next random pop culture question needs to be. It’s critical that we find out whether white&gold or blue&black identification correlates with voting intention.

  2. Mike N

    I believe the expression is lol!

  3. Auntie’s politics page is currently linking to an interesting page at May 2015 with multiple forecastings of May results, all predicting a NOC. The Ladbroke’s one isn’t fully explained – presumably it’s their “evens” offer.

    Seat predictions as CSV are:

    Party, May2015, EF, E&c, Grauniad, Ladbrokes
    Con, 270, 285, 279, 276, 280.5
    Lab, 271, 279, 283, 271, 272,5
    LD, 26, 25, 23, 27, 28.5
    UKIP, 4, 1, 3, 3, 5.5
    SNP, 56, 39, 40, 51, 38.5
    Grn, 1, 2, 1, 1, 1
    Other, 22, 20, 21, 21, 21.5

    See http://may2015.com/featured/five-election-forecasts-concur-labour-tories-to-fall-at-least-40-seats-short-of-a-majority/

  4. Interestingly Balamory is split between two constituencies.

    Most of the villagers are on Lid Dem held Mull in the 4 way marginal of Argyll & Bute, albeit with a view across the Sound to Ardnamurchan which is Charlie Kennedy Territory.

    Meanwhile Archie is holed up in his castle – Fenton Tower in East Lothian – a Labour vs SNP battleground.

    Early polling suggests:
    Miss Hoolie – SNP – educator and idealist
    Archie – Conservative – lives in a castle, obvious Tory
    Josie Jump – Lib Dem – believes in exercise and electoral reform
    Edie McCredie – Labour – old style Bennite trade unionist
    PC Plum – Green – University friend of Patrick Harvie
    Spencer – Republican – Religious right values voter
    Suzie Sweet – UKIP – petty borgeousie shop owner
    Penny Pocket – SSP – resents her oppressive boss Suzie

  5. @Andy

    You may be right that Green/LD crossover has occurred in some parts of the UK but not in others. This is certainly suggested by some of @Statgeek’s graphs .

    The problem with pursuing this is that England-only polls are few and far between and it would be almost impossible to gather the data needed to decide the matter.

    At 5.17 pm on Feb 8 I posted some suggestions on steps that might be taken to establish that crossover has occurred.

    No one seemed interested in engaging in debate about this or offering improved suggestions, but I think a prerequisite would be to have enough polling data to be able to establish whether two VIs are reliably different over a period as short as a fortnight or at least a month. The rarity of England-only polls makes this impossible, I think.

    Like you, I can’t see the LibDem VIs rising as much as the models predict but I am not sure that Ukip will notch up the fivefold increase you predict. It seems quite possible that some of the present support will melt away before the election.

  6. In slightly less important news, possible-Lib Dem-leader-on-May-8th Tim Farron has announced that the party has a duty to provide stable government, which I interpret to mean they’ll offer confidence and supply to whoever can put together a majority, regardless of who has most seats or most votes:

    http://www.newstatesman.com/politics/2015/02/exclusive-tim-farron-lib-dems-will-have-back-labour-if-they-win-more-seats-tories

  7. I think Balamory deserves its own thread

  8. NORTHUMBRIANSCOT

    LOL…..has to be analysis of the month. :-)

  9. MIKE N
    “I think Balamory deserves its own thread”
    ______

    It will certainly keep Jasper happy.

  10. Oh, best laugh of the week.

  11. @Barbazenzero

    Three of those forecasts put the Conservatives with the highest number of seats. And the other two would probably mean that the Conservatives would have the most votes.

    But

    Once one factors in SDLP. Green and PC every one of those forecasts puts Ed in Downing Street.

  12. Personally, I’m hoping Ashcroft will move on to the Night Garden after Balamorey. The Pontipines are obviously Labour and the Wattingers, Conservative. But what of the rest? Could be a marginal constituency with all to play for…

  13. What’ll be the result in the General Election.

    Wouldn’t you like to know.

  14. Balamory that is!

  15. NORTHUMBRIANSCOT
    Respect & genuine LOL

    LITTLE RED ROCK
    Once one factors in SDLP. Green and PC every one of those forecasts puts Ed in Downing Street.

    That’s how I read it, although it must be allowed that the NS & grauniad would say that, wouldn’t they.

    I wonder whether Dave & Ed will declare at the last minute that they’re both “new men” and agree to do a 50/50 job share for both themselves and their MPs with one month and on one month off, leaving opposition to all the “others”.

    I suspect they’d each find that much easier and more fun than negotiating anything at all with UKIP or the SNP.

    In much less than a five-year term you would truly find it hard to distinguish between them.

  16. Oh, for the days when politics was simple.

    A choice between the gibberish of Bill and Ben, with only the plaintive cries of Weed in the distance.

  17. @ OLDNAT

    There is still plenty gibberish for those who want it. And plaintive cries too.

  18. Naturally, YouGov have announced the people’s verdict on THE DRESS.

    Quite properly, for a Hebridean story, Sturgeon was the first party leader to announce official party policy that it was white & gold (provoking a furious Cabinet row with Roseanna Cunningham).

    Milkband’s late intervention is thought to indicate a willingness to accede to Sturgeon’s demands on everything – even though it puts him dangerously out of step with public opinion.

    https://yougov.co.uk/news/2015/02/27/its-official-dress-blue-and-black/

    74% blue and black, 16% white and gold

  19. @ Barbazenzero

    No need for an actual election then. The modelers have got it all sorted in advance.

    More seriously it is strange to see such convergence emerging from such disparate sets of calculations. Three of the models offer election-tomorrow snapshots. The other two build in substantial swingback/regression-to-mean between now and the election. And yet, armed with that information you probably couldn’t tell which was which from the projections May2015 (and now you) have reproduced. The EF team publish detailed seat projection lists for both types of model (current snapshot and May 7 projection). In their case the two respective seat tallies look markedly different. Why don’t the three snapshot projections look more like the EF snapshot projection and much less like there regression-to-mean adjusted May 7 projection?

    I suspect that the three snapshot-only models don’t have good algorithms for feeding in new polling data and that they are all rather badly off target. The only alternative is that it is the EF snapshot that is off target – a suggestion that doesn’t sit well with the evidence that it has consistently outshone its competitors in predicting new Ashcroft constituency results.

    We already know that May2015 has done a poor job of predicting Ashcroft constituency polls. But we are unable to pass any judgements on the others because they don’t publish the evidence that would allow critics to pull them apart.

  20. It’s white and gold. I don’t see any blue or black. What’s wrong with me!

  21. Surely someone is talking about the new TNS Scotland poll. Not a mention? In comparison with the last one, it is… interesting.

  22. Couper

    Typical party loyalist – just toeing the party line! :-)

    JR

    We have been. Comments are in the thread pages, because we always see the daytime released polls before Anthony gets a chance to post on them.

    Poor guy has to work for a living.

  23. This looks like an interesting piece of research in the Netherlands.

    We find that simple polling outcomes by themselves do not affect subsequent vote intentions. We do find evidence for a subtle but relevant bandwagon effect: An emphasis on growth in the opinion polls stimulates subsequent support. However, there is no evidence that the bandwagon effect is more apparent among people who were on the fence.

    Free to anyone with an academic login – sadly not me any more.

    http://ijpor.oxfordjournals.org/content/early/2015/02/19/ijpor.edu041.short?rss=1

  24. @ Unicorn

    I agree with you, but what I observe in Scotland is that outside of seats where LD holds the seat, according to Ashcroft, the LD vote has absolutely collapsed and they either tie with SGP or come behind them.

    The same pattern exists in England in terms of by-elections late in 2014. So my conclusion is that outside of the 46 seats in E & W, unless there is a very compelling reason to vote LD people will not and it is quite likely that Green could come ahead in a majority of those seats.

    The tricky question will come in seats like Norwich-South or Bristol-West where LD and Conservative voters will have a chance to prevent a Labour victory by voting Green.

    As for UKIP their vote now looks like it is spread fairly evenly across England and Wales, but that still might lead to some interesting three way fights in the Southeast and Midlands (including East Anglia).

  25. UNICORN

    The “interesting” seats that Ashcroft polled are more likely to be ones that do not fit the general polling trends.

    I think that statistical modelling of predicted seat totals is only going to be as good as the underlying assumptions (axioms) of the model, even if the statistical mathematics are sound. Not all of these are even stated (presumably all the models assume that there will not be a major financial crisis between now and the start of May for instance). However, I think it provides a useful pointer (and broadly matches a qualitative assessment that I have done of target seats).

    Personally, I my philosophical (epistemological) opinion is that trying to do statistical predictive modelling of individual seats is a fool’s errand, no matter the quality of mathematical modelling used as the data is nowhere near comprehensive nor reliable enough.

  26. Andy Shadrack

    I really can’t see many LD voters tactically voting for the Greens in Bristol West. They’re the incumbents!

  27. @OLDNAT

    Sorry then. I looked through several pages of comments and didn’t see any. I count on you lot to educate me, and you seem to be falling down on the job. ;)

  28. I hate to distract you from your poll-based prognostications, but Leonard Nimoy has died, a fact I find saddening.

  29. J. R. Tomlin

    Looking back, we haven’t actually said much about the TNS poll – possibly because it’s just another pollster, with only ancillary interest in political polling, falling into line with every other source of information.

  30. Martyn

    My Twitter timeline contains little else.

  31. @OLDNAT

    It isn’t really ‘just another poll’ though considering that it shows a 5% increase in SNP support since the last TNS poll. If this is rather remarkable considering the general consistency of Scottish polling.

    One explanation, of course, is that the last poll was merely an outlier or that TNS has changed its methodology. Or maybe the SNP has had a rather large jump in support. But just another poll? I’m not sure that’s true. :)

  32. Scottish voting intentions for the May 2015 general election (TNS)

    SNP 46% (+5)
    Labour 30% (-1)
    Conservatives 14% (-2)
    Greens 3% (-3)
    UKIP 3% (+1)
    Liberal Democrats 3% (-1)

    The SNP +5 is well outside the usual normal variance noise.

  33. @Oldnat

    I’m at work. Will talk later

  34. J. R. Tomlin

    TNS are an odd bunch.

    The fieldwork period for their last 2 polls overlapped (14 Jan – 2 Feb and 30 Jan – 22 Feb)

    Their Jan poll was out of line with every other pollster, so didn’t have that much credibility, unless they were uniquely discovering a trend unseen by others. This poll just returns them to the mainstream.

  35. UNICORN

    No need for an actual election then.

    Just a bit of fun, as Peter Snow used to say in his “swingometer” years.

    We already know that May2015 has done a poor job of predicting Ashcroft constituency polls. But we are unable to pass any judgements on the others because they don’t publish the evidence that would allow critics to pull them apart.

    Fair comment. I’m not defending that site but merely intrigued that they have such close current projections despite their disparate approaches. I gave up pretending to be a statistician when I got my degree in stats in ’71 and returned to my previous employment of computing in general and symbolic logic in particular which has intrigued me ever since.

    I’ve never understood the supposed XOR between Lab & Con when they share their key policy of maintaining the establishment.

  36. Not unexpectedly, Jim Murphy has confirmed to his CLP that he will stand for Westminster in May,

    Herald cheekily headlines the story as “It’s a Yes: Jim Murphy will stand again in East Renfrewshire for Westminster”

  37. OLDNAT

    I am a different kind of nerd in that reading that headline I thought of Jim Hacker.

    Sad that Leonard Nimoy won’t be coming back like Spock did. RIP,

  38. J. R. Tomlin

    Though these are strange times.

    Pollster says 46% SNP for Westminster. We shrug our shoulders and say “Meh. What’s new?”

  39. @ Funtypippin

    I encourage you to look at the Bristol local government elections results for 2013 and 2014 in Bristol West:

    http://www.bristol.gov.uk/page/council-and-democracy/election-results

    The City of Bristol website provides a pdf map showing which City wards fall within Bristol West and also provides results by ward by year.

    My observation is that LD voters and others have already moved over to Green in that seat, much like what happened in Brighton-Pavillion.

  40. @OLDNAT

    Strange times indeed!

    What I wonder is if TNS changed their methodology in view of their out-of-line results last time. I was hoping someone here might know.

  41. Kudos to Northumbrian Scot although would PC Plum really be a Green?

  42. @J R Tomlinson

    We look at all the crossbreaks. So the poll is not a surprise, if a poll completely out of kilter with the crossbreaks came along then that would spark discussion. I don’t like TNS much because they give a combined score for the VI & squeeze question and they have large numbers of don’t know/won’t vote/refuse

    The effect it will have is to dent LiS morale. I am detecting a bit of panic in the LiS ranks

    @Old Nat

    Not a surprise, I’ve long said, Murphy won’t give up Westminster.

  43. @J R Tomlinson

    No there’s no change in methodology & TNS has a pro-Labour house effect.

  44. @Mikey

    If crime is the problem, more PC Plums are the answer.

    A one man crime fighting machine. He could finish ISIS off before his porridge in a morning.

  45. J. R. Tomlin

    No obvious sign of changed methodology.

    TNS have only done 2 post-referendum polls, but I’ve just spotted one intriguing point.

    In their Jan poll, they had the gender gap for the SNP closed – with M 40% : F 39%.

    This time it is M 44% : F 48%

    I might have to do a bit of digging into that. Just as well you dropped by! -)

  46. Gosh, some people here must have a LiS “panic” detector (also know as: A risible, partisan comment generator).

  47. Main methodological change for TNS is sampling voters who don’t live in the Stockbridge Colonies this time.

    Green sub sample in Lothians fell to 0% from 21%. Andrew Agnew was obviously out when they called round. ;-)

  48. That dress.

    The same picture, on the same laptop, from the same seated position:

    I see it as White & Gold; my son sees it as Blue & Black. I didn’t even tell him that it was up for debate. I just asked him: “What colour is that dress?” He immediately replied: “Blue”. What about the trim? “Black, or maybe Charcoal”.

  49. Just chatting to myself beside the playground fence ….

    That dress

    My son in the USA posted “in our house alone I see white and gold, [my grandson] sees blue and green.”

  50. @COUPER2802

    It is however very much out of kilter with their last poll and both were a full poll, not a crossbreak.

    The difference is outside normal margin of error noise, which means that it could (note I said *could*) be the beginning of showing an increase in the SNP lead because a +5 from their last poll is remarkable if there was no change in methodology.Or of course, it could be totally meaningless. We’ll see but I was surprise that no one here remarked on some changes that I see as raising questions.

    As OldNat said, strange days. :)

1 5 6 7 8