There have been several polls this week showing both the Conservatives and Labour up at 35%, and given it’s a zero sum game that suggests the other parties are getting squeezed. There is a general expectation of this sort of squeeze as the election approaches – the nature of First Past the Post is that votes for smaller parties don’t stand much chance of being translated into MPs unless they are geographically concentrated and as the election approaches the media coverage almost inevitably focuses ever more upon the main contenders (though more on that later).

The graph below shows the average poll scores for the Greens and UKIP across the nine regular pollsters (for those pollsters who do several polls a month, I’ve taken their average across the month).


UKIP have been on a slow but steady downwards trend since their support peaked after Douglas Carswell and Mark Reckless’s defections. This last month the decline may have flattened out a bit, but that is largely due to MORI having an unusually low score for them last month that jumped back up this month – without that the line would show a smoother downwards trend. The Green party’s advance seems to have halted last month and started to fade a little this month.

Of course, just because they might be getting squeezed shouldn’t distract from the fact that UKIP and the Greens are still doing incredibly well compared to the last election. In 2010 UKIP got 3%, the Greens 1% – both parties could suffer a bit more squeezing and still end up quadrupling the vote they got last time. For UKIP, there is also good news on the horizon, next week the campaign broadcasting restrictions kick in, guaranteeing them coverage as a major party. For most of the last few months UKIP’s media coverage has largely consisted of the latest row or resignation for inappropriate comments. Next week the broadcasters will have to start giving them more neutral coverage alongside the Conservatives, Labour and Lib Dems.

In other news, we had the monthly Survation poll for the Daily Mirror out today. Their topline figures were CON 32%, LAB 33%, LDEM 8%, UKIP 18%, GRN 4% (tabs.

384 Responses to “Are UKIP and the Greens getting squeezed?”

1 2 3 8
  1. Anthony

    How do you give “more neutral coverage” to another UKIP candidate resigning because the party treats a racist, Islamophobic comment from an MEP as a “joke”?

  2. Thanks AW

  3. @Andy S (FPT)


    So your graph is very interesting, but how much of it, if the above is an accurate reflection of a YouGov “house effect”, is attributable to “swingback” methodology being imposed on the data.

    If Kellner and other pollsters and forecast projection technicians believe in “swingback” and model for swingback how do you differentiate between “effect” and reality.

    A very long time ago I did a soil sample test in Wales for the National Farmer’s Weekly, who were having spontaneous abortion occurriung in some Finnish Ewes. Until I came back with lab results no one had a clue why it was happening, but when they top dressed the field with the trace element I found missing in the soil sample the abortion cleared up.

    So when I see all these polls I simply ask is an unweighted LD value of .289 reasonable under the circumstances and is a weighted value of .305 Kellner imposing his “swingback” theory of the respomdent value?

    Stats are very funny things and if you “massage” them long enough they can do anything.

    Where do I begin?

    I’ll keep it short because I’m full of cold and very tired.


    Part 1

    CUSUM is a time-weighted SPC chart that analyses trends (subtle, below 1.5 sigma) in time sequenced data up to the current time. It’s not a regression on past data that then try to project it forward. Personally, I don’t regression stuff. That is Unicorn’s speciality.

    It can only work on the data provided. You can’t validate a daily poll with some ‘reality’ most of the time, unless a poll is carried at the same time as an election to compare against, as that ‘reality’ is unknown.

    CUSUM is so subtle that the different companies that poll can’t be mixed, as the noise introduced would dwarf the signal you are trying to read. You could run this system with other Pollsters, but they don’t have the quantity of data to be as good, in my view.

    Every Pollster has some house effects, but the analysis of a company’s data still provides a reference measurement to trends. Psephology doesn’t have precise alternatives to measure public opinion, apart from elections. Even then, a partial Council election cycle give you issues, as you can’t sure the seats/wards voting are representative. of the whole nation. (This is Rawlings and Thresher territory).

    There is no ‘massaging’ of the data. It creates a mean for the period, and then plots the cumulative sum of each data point in time sequence vs the mean. If just MOE is present, the CUSUM bobs up and down randomly.

    When the cumulative sum goes up or down in prolonged zones, it demonstrates changes beyond MOE. You can then measure the data points in one ‘zone’ (normally dicated by prolongued peaks and troughs)vs an adjacent zone. Use a T Test, Kruksal-Wallis, whatever best suits the data. You can look at CI intervals, p values, whatever.

    Take the UKIP claim (no, it’s more than a claim, it’s a provable statistical result) that UKIP have declined. They have – compare the rolling average, the first data points of the year vs the most recent, there is a difference. They had 8 consecutive YG polls below there 2015 mean. Seeing as this is between 14 and 15, even if the 14 from today was substituted for the last night’s 12, it would not alter the trend. The great thing with CUSUM is that has this and trend element, and doesn’t dump the data into an arbitary Month A or B

    If they were polling the same all year, the chance of eight data points being below the mean is 0.5^8 or 0.39%.

    So either UKIP VI has changed or a lucky 0.39% event has happened.

    What’s your guess?

  4. Anthony, many thanks for this new post.
    It will be fascinating to see how the ‘Big Three’ parties pick up the voters coming away from UKIP and Green

  5. Part 2

    Finally, here are the five poll rolling averages for 2015 (UKIP VI), up to last night:


    They are recording their worst averages for the year (or it’s a really freaky unlucky spell, all down to MOE.)

    Roger Mexico, Unicorn and myself try to ensure what we publish is fair, non-partisan and accurate, as do all the other people who visit UKPR. I use CUSUM because it does a really good job. I use CUSUM, even though I’m a Green, and quite frankly it’s all doom and gloom on that front. I still publish as it’s the right thing to do to be accurate and unbiased. I do not report biased data, never have and never will. I do not knowingly massage data. Data is what it is, and that gets reported. If I make a mistake, I take the sound advice the really, really top draw statisticians that give helpful advice and commentry

    CUSUM is proven over vast decades in industry in process control methodology. If you are unsure, there are lots of sites out there that explain SPC statistical tools. I only scratch the surface.

    If you would like see my method and any flaws you might think is there, let me send a link to my spreadsheet and calculations. I’m happy to share


  6. @AW

    Thank you.

    Very timely :-)

  7. Thanks Anthony.

    I agree that the purple graph is going down over that period.


  8. I would to love to know why we have seen a UKIP decline, maybe voters don’t want a Labour PM and are returning to the Tories.

  9. And would expect the Greens to be squeezed further, ending up with no more than 2-3%. But that would still be the best result eve.r. And they they will get at least a few second places and possible another MP. (In 2010 the Green vote went down but as Caroline was elected the whole party agreed it was the most successful election ever. And who would disagree?)

    They also have gained thousands of activists which should translate into better campaigning in future elections and they are, for the foreseeable future, a leftward drag on Labour policy.

    And if, at it is likely, the SNP holds the balance of power, then the Green Party’s influence will also rise – because you can expect the SNP, Plaid and Greens will work together. There would complete freedomfor each party to vote in accordance with its own policies, but just like in Europe in practice you can expect close cooperation and convergence.

  10. ChrisLane1945

    “It will be fascinating to see how the ‘Big Three’ parties pick up the voters coming away from UKIP and Green.”

    I think we know how the third of the “Big Three” pick up voters from UKIP and Green.

    SGP supporters will largely vote SNP this May, while UKIP supporters in Scotland won’t.

  11. The first opinion poll I was ever in.

  12. OLD NAT.
    Thanks for that; no wooden spoon then for those particular men in blue.

  13. Oldnat,

    Very good (and true)!

    But it is funny to see all this ‘waily waily waily’ abou the SNP’s rise, and the in particular the confusion, given that anyone who has followed the SNP government and the independence campaign (and indeed any statment von Surgeon or Salmond since) should have no problems whatsoever to work out how the do in May and what they do afterwards….

  14. @John J

    I think it’s quite simple.

    UKIP score big o two issues – immigration and Europe. I know they have more policies, but given where we are now, the election is about health, education, jobs, the economy etc. That move people to the usual suspects.

    The Greens suffer with this too. Once the big debate are on more convential topics, they struggle to get heard.

    That’s life under the FPTP system and related media.

  15. Chris Lane1945

    Don’t be sexist! Our Rugby ladies got the wooden spoon as well.

    Very useful implements, wooden spoons. Since we do so well at collecting them, the Scots term “spurtle” should probably be used.

  16. Plenty of scope for more squeezing. As late as December 2013, UKIP were getting 7, 8, 9% in the polls.

  17. OLD NAT.
    Very good; but the word men/man means human being as well in Latin.
    Do you think Labour will have a swing back to itself by May 7?

  18. Not sure why UKIP are classed as on a par with the big two (three) and the likes of the Greens are excluded.

    But the decision has been taken and so they are a “major party”.

    Time will tell. No.. this is no time for fence sitting, time will show that decision to be wrong. UKIP are not and never will be a “major party”.

    On verra.

  19. UKIP with ‘inappropriate’ comments? Come on Anthony racist, sexist and fraud are not ‘inappropriate’ comments!!

  20. For what it’s worth, UKIP are back where they were in March 2014 in Scotland –

    (Ok, so they are up from 3.3% to 3.4% – Big diff)

  21. I heard a commentator on Scotland arguing that the apparent unease of English people at the rise of the SNP give the SNP an aura of potency, that adds to their appeal among ordinary Scottish people.

  22. “Come on Anthony racist, sexist and fraud are not ‘inappropriate’ comments!!”

    They are not inappropriate?

  23. Crikey, I see what you all mean now. In that Survation poll, UKIP have crashed by 1%, plunging to the hitherto unplumbed depths of 18%.

    That’s some squeeze going on there with the Tories, on 32%, opening up a gargantuan gap of 14% over Farage’s steadily sinking souffle.

    At this rate, it’s quite possible that UKIP will receive no votes at all on May 7th.

    The Tories being the main beneficiaries of course.


  24. Chris Lane 1945

    “Do you think Labour will have a swing back to itself by May 7?”

    I’ll need to consult my Ouija board, and that could be a problem if the DUP make the regulation of their sale a condition of C&S –

    Mr Gregory Campbell (East Londonderry) – “To ask the Secretary of State for Business, Innovation and Skills if he will bring forward legislative proposals for regulating the sale of ouija boards” (Hansard).

    Without them, how can UKPR survive?

  25. CrossBat
    Well put. i was going to remark on the 18% myself but you’ve done it for me.

  26. I forgot how bad the moderation is on here sometimes. I wont be voting UKIP, but it’s open season on the comments against them on here from the resident left wing. Any slightly risky comment againt Lab and you’re straight in moderation!

  27. @crossbat @pete b

    Before you get too hung up on that 18% from Survation,

  28. PETE B

    Perhaps you should read this again :-

    From AW’s ” All about House Effects”
    29 Jan. 2015

  29. Rich

    Is rejection of racism and Islamophobia a “left wing” position?

    I’d have thought it more a mark of decency.

  30. @colin

    Gold stars for us

  31. OMNI


  32. Hardly open season Rich.
    Having said that UKIP’s share of the vote is still way higher than it was this time five years ago and if they can hold steady at around 13% of the share of the vote it represents a considerable success for them.

  33. @rich

    If I sound satisfied that UKIP are declining it’s not partisan, it’s because they’re following the trend that I predicted. I will also be thoroughly satisfied at the Greens declining for the same reason.

  34. re UKIP being a major party – they did win the European Elections last year. That’s got to count for something towards their “major party” status.

  35. Look, I know I’m a “UKIP decline denier” (sic), but, as Anthony himself says, we’re just entering an election campaign where Farage and his party will be back in the spotlight again. I suspect he’ll perform well in the TV debates, slinging tasty soundbites around, gunning for Cameron particularly, and his party will also benefit from what is bound to be slowly developing mainstream party fatigue after three weeks of inane slug-festing between the Tories and Labour.

    It’s worth remembering also that UKIP aren’t some rootless electoral bubble just waiting to be popped, they’re a well organised and funded political campaigning force, embedded in hundreds of constituencies throughout the country. They’ve maintained mid-teens ratings in most polls, sometimes higher, for very nearly three years now. They have MPs, battle-hardened campaigners and scores of councillors and MEPs. They believe they’re a party who’s time has come and they go into this election having won the last national one only 12 months ago.

    They just ain’t going anywhere, I tell you. 10-12% VI score on May 7th and mayhem-merchants in many a marginal constituency. Of course they’re not going to get anywhere near some of their heady polling spikes of 12 months ago, and they’re starting to drift very slowly downwards, but let’s return to that Survation poll today. Here we are, 40 odd days to go to the election, and a reputable political pollster has them on 18%.

    18%, for pity’s sake!!!

    If that doesn’t frighten the life out the Tory and Labour camps then they’re in the wrong game.

  36. @Keith P

    Indeed. It was one of the factors Ofcom too into account.

  37. @Crossbat

    Agreed. I guess we’ll find out pretty soon either way. The short campaign starts on Monday.

  38. @Omnishambles

    Ah, yes, our trusty old house effects yarn. Forget Survation because they reside in a virtual political world of their own making.

    Daily Mail and Daily Mirror both consulting their lawyers, I suspect!


  39. If Bercow is dumped as Speaker before Monday, does he have to seek nomination as a Tory? candidate to hope to be re-elected?

    Presumably, the other Westminster parties would put up candidates against him, if the local Tory party did select him?

    With no Speaker in place when the UK HoC resumes, presumably the first action has to be to appoint one – taking one off another party’s tally.

  40. Yes thank you, I did read and understand the article about House Effects. It just seemed amusing that after an article suggesting that UKIP’s VI was slipping downwards we suddenly get an 18% for them.

    Are all these ‘racist’ and ‘islamophobic’ references about David Coburn’s recent remark? If so, I can’t see why it is any different to Vince Cable comparing Brown to Stalin. It’s just a joke. Are we not allowed to joke if one of the people in the joke is brown?

  41. Ukip up at 18% in the latest poll published by the Daily Mirror. With a lot of media time guaranteed, perhaps predictions of their imminent demise are premature.

  42. I am still in admiration of Salmond’s cunning game, antogonise middle England with so many threats, that DC will win, or be as close to as needed, and then turn round to SCotland to point to a Tory government, not elected by “us” so the SNP 2016 manifesto for SNP in Holyrood has another referendum on it, he just wants an Independent Scotland with himself as the first President, no interest whatsoever with the livelihoods of his constituents.
    Should we not just get on with it? It was Blair’s mistake, SNP are now in a win, win, win scenario. Independent Scotland, get on with it and good luck!

  43. Anyone been called by GKM marketing (or similar) doing a general election survey?

  44. David in Oxford

    I’d normally recommend a mild sedative, but you may require something stronger. :-)

  45. @oldnat

    I have teased my Tory parents with the suggestion that, since the two big parties may not wish to give up an MP if the arithmetic is tight, perhaps you could end up with a SNP member as Speaker? Maybe even A. Salmond?

  46. @ David in France,

    Not sure why UKIP are classed as on a par with the big two (three) and the likes of the Greens are excluded.

    Because Ukip came first in a national election and the best the Greens have ever managed is third.

    @ Rich,

    Lol. Where exactly do you think Anthony’s allegiances lie?

  47. @David in Oxford

    We’ll be hearing a lot more from Nicola Sturgeon in the debates. She happens to be the leader of the SNP. Unlike Salmond.

  48. @crossbat
    [your impassioned monologue]

    Look, I know it’s hard… but let it go. The longer you wait, the tougher you’ll find it. You will always have memories of polls gone by. Opinium – gave UKIP 20% in January. Our old friend Survation – 25% last year!

    Those were happy times. Happy polls. That’s how you should remember UKIP, not desperately clinging onto double digits as the last debate bounce fades away and campaigning intensifies.

    Good luck, I know you will find a way to cope.

  49. @rich
    The site has gone downhill, as we have got closer to an election. Don’t expect that to improve. Too many are too emotionally involved.

  50. @ Pete B,

    I can’t see why it is any different to Vince Cable comparing Brown to Stalin.

    Because the basis of Cable’s (nasty and unwarranted, but what else can we expect) comparison was Brown’s actions, and the basis of Coburn’s comparison was Yousaf’s name.

1 2 3 8