Today’s twice weekly Populus poll has topline figures of CON 33%, LAB 38%, LDEM 9%, UKIP 13%. Full tabs are here.

Meanwhile this morning’s YouGov poll for the Sun had topline figures of CON 34%, LAB 39%, LDEM 8%, UKIP 12%, full tabs here. Given it’s the last day of the month we can look at YouGov’s averages for the whole of February, which gives us figures of CON 33.2%(33.0% in Jan), LAB 38.9%(38.7% in Jan), LDEM 9.3%(9.2% in Jan), UKIP 11.8%(12.5% in Jan), so no real movement month-on-month.

164 Responses to “Latest Populus and YouGov figures”

1 2 3 4
  1. First?

  2. Oh polldrums!

  3. I dunno, UKIP dropping nearly a whole percentage point could be significant, although we’ll have to see whether it continues.

    A small amount of good news for both parties’ main supporters, but as you say nothing to get too excited about.

  4. Polldrum-ish, with the caveat above. Maybe the by-election did stop the UKIP momentum?

    I do begin to wonder what would have to happen to change the polls now. We have had austerity, scandals of various sorts, vote-rigging allegations, and although Statgeek’s regional figures show some movement overall polls are fairly stable The world situation is pretty volatile, but I would not wish a war on anyone.

    Perhaps if there was a shortage of tea-bags something might happen?

  5. It is indeed unexciting but that’s where the polling ‘news’ for 2014 lies so far:

    1 Slow reduction in the Labour lead as perceptions of improvement in economy grow : Not Happening.

    2 Romanians And Bulgarians : No Effect

    3 By-election: No Effect,no new ones scheduled.

    4 Flooding: No Effect.

    5 Toxic Swamp (to purloin Alec’s handy and appropriate term) : No Effect

    It looks to me more than likely it’ll stay this way until the budget and I won’t presume to guess both what will be in that and its effect. – and I shall certainly be hoping that no-one generates any vote influencing excitement over the Ukraine matter.

    As for UKIP,their moment comes in May.I’m not as convinced as some people are that UKIP will win but I seriously underestimated them last year and there’s plenty of time yet.

  6. Just seen someones tweet where he said he wanted to vote Green but had no candidate, didn’t like the main parties so he would vote UKIP. This says quite a lot about the UKIP support.

  7. Farage’s address to the UKIP Torquay assembly well crafted; immigration is the big item and the new clincher that it is not Bulgarians and Romanians from Bulgaria and Romania but the 1.5 million of them fleeing the collapsed economies of southern Europe who will now come to Britain. “If you want an immigration policy like Australia’s vote UKIP!”. says

  8. Farage’s address to the UKIP Torquay assembly well crafted; immigration is the big item and the new clincher that it is not Bulgarians and Romanians from Bulgaria and Romania but the 1.5 million of them fleeing the collapsed economies of southern Europe who will now come to Britain. “If you want an immigration policy like Australia’s vote UKIP!”. says Farage.

  9. Sorry – shaking finger of doom.

  10. Been a long time since I’ve done Lefty’s Graph of the Week, but I’m going to try to get back into it. Here’s a nice one that has something for everyone (well, something for Lab and Con supporters anyway, and possibly the smaller parties by extension).

    It’s a graph showing:

    On the y-axis – the difference in vote between the governing party and the main opposition at each GE since 1955

    On the x-axis – the gap in the polls between the Govt and the main opposition party 12 months before the GE.

    Basically, it shows how much Govts catch up over the last 12 months. I’ve produced it because I was fed up with hearing idle “x% ahead is/isn’t enough at this stage to do y at the Election”.

    What about that then eh? There are two great outliers (79 where the WoD changed everything and 87 where I guess the great big boom and tax giveaway in the run-up to the GE changed everything). Apart from those, there is a clear and consistent theme which raises 2 very important points.

    1) There is clear evidence of a closing of the gap in the run up to elections BOTH ways. Govts that are unpopular catch up, ones that are wildly popular don’t keep that going.


    2) There is a bias.If it were just as simple as that, the trend would go through the origin. It doesn’t. A linear trend line hits y=0 at x=-7. A parabolic fit cuts at x=-9 ( values are -6 and -8 if we ignore the 79 and 87 outliers).

    This means that a Govt which is ~6 to 9% behind in the polls 12 months before an Election would normally expect to make up that ground and tie the GE in terms of vote share.

    Take the lesson from that which you find most fitting. To me, it suggests that all possibilities for GE15 are still on.

  11. My gut feeling with Nigel Farage is he’s going to cause some serious political damage in May. A lot of people I know from all parties, including some card carrying Labour supporters, are going to vote UKIP for the European elections. I’m a LD who is really concerned about the social implications for us all if we don’t get a proper grip on immigration. I’m seriously considering voting UKIP myself, something I never thought I’d say.

  12. Lefty –

    Steve Fisher’s election forecast thing (which I keep meaning to write something about) is essentially that – the best possible attempt to project an election result in May 2015 based on polls now and how the polls have tended to move in the run up to previous elections:

    Make it if what you will, but for what it is (which beyond all else, is just a simple application of “here, ignoring all other factors, is what would happen based purely on current polling and the way polls have normally moved in the past”) it seems to be extremely robust. Equally, extremely important to note the huge error margins!

  13. Achh…error. I had the 83 and 87 election results the wrong way round. This is how it should be.

    The relation is even more linear than before and the y=0 points are at ~x=-6 to -8.

  14. Lefty – re 86/87 the Government took a big hit in the polls early in 1986 from the Westland Affair and whilst it unwound there was still plenty of impact in the polls 12 months before the GE.

    We have to remember that gaps were larger in polls pre-97 because the polling companies were not as sophisticated so it is difficult to interpret.

  15. AW.

    I noticed when I did this that if you plot the GE-15months data on the x-axis, the relationship is nowhere near as clear. And I suppose that SF’s approach is a purer one than mine, in that he’d accept 79 and 87 as part of the data set with the resulting big spread in his 95% CIs. The pragmatist in me would say that those were exceptional, unlikely-to-be-repeated circumstances, and that barring them, the relationship at GE-12 months is quite astonishingly consistent.

  16. Good Evening All.

    AW, JJ and Lefty L.
    I remember Labour’s hopes in 1987
    Labour won Fulham in a by election and then got smashed in the 1987 GE, not long after.

  17. If I remember previous figures correctly, Populus this time shows a sharp drop in SNP with a large increase in DKs in Scotland (presumably people willing to consider voting Labour).

    Now, what would a Scottish Labour ‘revival’ blocking Lefty’s possible Tory resurgence do to GB politics – if anything – particularly if coupled with a Douglas Alexander style Devo-Max programme?

  18. On the other hand, YOUGUV show Labour and SNP almost level pegging.

    UKIP seems still to have little effect on politics north of the Border – indeed, the whole UKIP thing seems to be passing us by entirely. They seem to represent all that’s worst about the ‘Little England’ mentality which many Scots find so irritating. It will be interesting to see whether they pick up a few votes in Scotland at the Euros.

  19. @Lefty

    Thanks for that analysis.

    Might you consider doing an alternative version, showing the position of the polls 3 years 10 months after the date of the previous general election (i.e. just about now in the current cycle), compared to the outcome of the subsequent GE? I’d find that more a bit more meaningful.

    As it stands, there is potential confusion of cause and effect, because the date of the GE was itself dependent on the polls. In elections where the Government chose to hold a GE after 4 years, the polls by and large had moved decisively in their direction over the previous 12 months. In those which went the full 5 years, the Government was generally in trouble and held back.

  20. I must admit that I tend to reject this kind of approach more or less out of hand for two reasons:

    First because the samples – elections- are so small and made smaller if you then reject anomalies.

    Second because for such trends to be real you have to believe that vast numbers of ever-changing voters are governed in some mysterious way by influences on their voting behaviour regardless of circumstances,events,issues, personalities etc.

    I call such claims ‘Norwegians’ as in “Norway have never beaten England at football” – they’re statements that are only true until they’re not true,not some innate law of football or politics.

    It’s sensible to point out to anyone who sees Labour 6% ahead and says ‘Game Over’ that history shows it can be otherwise but that’s as far it goes.

  21. As for the immigration ‘problem’, apart from a surge in Polish folk five years back we’ve really not noticed much immigration here except in certain specific areas which were targeted by the government for ‘asylum seekers’.

    I always laugh when I hear people say that the UK is ‘overcrowded’. Exactly which part of Dumfries and Galloway, the Borders, Argyll, Perthshire, Inverness-shire, Ross, Cromarty, Sutherland or Caithness (which together make up about 30% of the GB land area, but with around 1.5% of the population), or large parts of Wales, Northumberland & co. are they taking about? Lets’ get this clear: the UK is not overcrowded. only certain parts of it are – probably less than 30%.

    But that’s enough from me.

  22. @LeftyLampton – I think that with the rise of four-party politics this is very likely to be an exceptional election too.

  23. Well, enough from me on UKip etc.

    But CL1945’s point is important. People will vote in a by-election in ways which they would not do in a GE. A protest vote which doesn’t actually threaten a government’s majority is quite different from deciding that the main opposition party is fit for government.

    But at the moment do either of the two big parties look like majority government material?

  24. Well if Nigel Farage is to be believed* then come June 2015 then three of the four main parties will be probably be looking for new leaders – whichever of Ed M or the David C isn’t the new PM, the Lib Dems unless something unlikely happens, and Farage himself, as UKIP are not going to win any seats and he’s now promised to resign if/when that happens.

    *I don’t.

  25. Sen5C

    I think you are misunderstanding my reason for doing this. I’m not trying to make a scientific prediction. I’m looking for general trends that might give pragmatic guidance (it’s the difference between looking at data as a scientist, or as an engineer).

    I look at that data and I see that there is a trend that is so clear, it probably helps to give some guidance. The Norwegian argument only applies if you look at that data and try to draw definitive conclusions from it. If you try to do that, you end up with Steve Fisher’s approach which tells us that we can be reasonably confident that Labour will gain between 181 and 370 seats at the next Election.

    As an engineer, when making first-order assessments, I tend not to worry too much about 95% confidence intervals. I look at trends and assess whether they give me a guide as to likely outcomes. What I’m looking for is something to help me make up my mind when someone puts a gun to my head and says, “You don’t have all the information, and you’re dealing with a messy problem, but you now must make a decision and you don’t have the luxury of dealing with big confidence intervals.”

    In this case, with a gun at my head to make a decision, I’d say we’re likely to see a closing of the polls that puts us somewhere in the range of very small Lab majority to just about level pegging.

  26. AW

    I shall be very interested in your article on Fisher. I cannot see any real logic in his approach but perhaps history does repeat itself that accurately.

    FLT -excuse my stupidity. I can only claim that I did look up Dnipro on Google Earth and got a riverside cafe in Kiev (the River Dnieper) and there was a suburb over the other side of the river also so-named. So I assumed it was like ‘West Ham’, say!

    But I missed your reference to Birmingham and Edinburgh, silly me.

  27. Chris M.

    Maybe. But then again, the data in that graph covers periods of both 2 party and 3 party politics, and both fall on the same general trend line.

  28. @Chris Martin

    If the Lib Dems recover something of their previous %s and UKIP don’t drop below 13% there could be absolute carnage at the GE.

    On the other hand…….

    And there’s still a long way to go, as Lefty was pointing out.

    Although Labour’s present % lead give them the prospect of an overall majority, the simple Swingometer model cannot give us a complete understanding of the implications for GE of the present Polldrums. The point of doldrums is that nothing is happening. No-one can tell which direction the wind will be blowing in when it arrives…

  29. ….and have UKIP seriously made Neil Hamilton their campaign manager? The man who lost a majority of over 16,000 in one of the safest seats in the country, by over 11,000 votes?

    Good grief.

  30. And as I mentioned a couple of days back, if the GE does turn out to be level pegging, then the Tories will really rue the day they decided that ‘saving’ the House of Lords was a greater priority than boundary changes…..

  31. Re: Pre-predicting VI movements in last 12 months of a parliament.

    I suggest that we are in totally new and uncharted territory here as Prime Ministers can no longer time GEs according to “Best time to go”. It is much harder for an incumbent party or coalition to “time the economy” just right for maximum benefit when you have a fixed term parliament. Therefore the historic spread of VI behaviour is based on different criterea as it is date based on a chosen and desired end date rather than a fixed target date of the GE as now.

    Can we therefore deduce much, even by a mythicl “law of avergae behaviours”? I doubt it!

    However, that doesn’t mean that it won’t fall within the range that incidentally coincides with the data on past elections. We simply do not know.

    My hunch, and that is all it is, is that it will be pretty darn close!!

  32. sorry! Average – not avergae!

  33. AW & LEFTY

    Thanks both.

    LEFTY- I don’t pretend to understand the finer points of statistical theory underlying your data.
    But does your graph allow for the seemingly permanent schism in the LD vote, soon after the 2010 GE-and the equally permanent addition to Labour’s VI ?


  34. Apparently UKIP want to have ‘overseas territories’ MPs like France have. UK’s overseas territories have a total population of 350,000, so around 275,000 potential electors, or about four seats. Not sure how they’d be divided, how they’d travel all over the world to meet with constituents or how they’d deal with the West Lothian question made even more serious!

  35. Neil Hamilton as UKIP’s Campaign manager – you couldn’t make it up!
    I also read an article in the Torygraph the other day urging Cameron to put Michael Gove at the heart of the Tories election campaign.

    Yes UKIP and Tories – please give Neil Hamilton and Michael Gove as much prominence as possible!

  36. @ John B

    The “Little England mentality” soundbyte is getting a little tired, a phrase constantly thrown by libertarians the moment someone with a contrary view sticks their head above the parapet. Why for heavens sake do you seem to think a big increase in our population is something to be wished for? Is the solution concreting, bricking and tarmacing parts of the country where there’s little infrastructure in place? Where do we stop, 200 million?

  37. Looking at the samples in the Populus poll, they look very ‘sensible’ not requiring much weighting in any category. The split between public sector employee VI and private sector VI is very great indeed. Likewise the North / South VI divide is huge.

    I don’t know why Populus lump in Wales with the SW whereas YG IIRC lump them in with Midlands.

  38. so many dinosaurs on this website! As if what happened in a general election in 1987 is of the slightest relevance to a general election in 2015. It would be like people in 1974 looking at what happened in 1945. It’s bonkers!

    UKIP, the implosion of the liberal democrat/sdp centre vote, the first full blown coalition since the 1930s are all material differences that have occurred in the last 4 years, but people still think we’re in the 80s or early 90s! unbelievable.

    The 2015 election will be utterly unlike ’92, ’87 or any other general election that 50 year old + grey beards talk about.

  39. “4 Flooding: No Effect.”

    Actually there were some residual damp problems.

    Disappointing not to be getting more updates on how ORFUL things are in Ukraine as I would never have knowed otherwise.

  40. Thank you for that Peter C. Most helpful.

    Funny thing is spadger, 2005 was very similar to 1959. Who’d have thought it, eh?

  41. One point that should be noted when making comparisons based on trends in previous elections, is that in 1987 when the Tories had a 101 seat majority, on current electoral boundaries, with the votes cast in exactly the same way, their majority would be just 20.

    And with Ukip disproportionately taking former Con votes, left leaning Lib Dems forsaking the party and an increased BME demographic leaning towards Labour, they have a mountain the size of Everest to climb.

  42. I have to say that the shift in the UKIP VI is probably more significant than margin of error given that it is an average over the month.

    I don’t know if Farrage has been around less in February and that’s what’s done it. If he has been around the same amount, especially as the Euros are coming, perhaps we are seeing people seeing UKIP as less credible when it comes to an actual election.

    Still only speculation though; most probably linked to prominence of profile IMO.

  43. @ Lefty,

    Been a long time since I’ve done Lefty’s Graph of the Week, but I’m going to try to get back into it.

    Hurrah! I’ve missed them.

    And that is a fascinatingly linear graph you’ve got there. I agree with Phil that it might be worth doing some with “X months after last general election” to control for opportunistic election timing, but even allowing for that the tightness of that correlation is remarkable.

    2005 was very similar to 1959. Who’d have thought it, eh?

    In both elections the main movement was a swing from Labour to the Liberals, so they may have more in common than we generally think…


    Totally agree…. 2015 is quite special. Coalition Governemnt coming in after 5 years (when did that last happen – ’45?). Collapse in LD support from strong showing previously. UKIP dividing the right. Europe tearing the tory aprty apart. Ed M doing his own thing and moving to the left………. It will be a cracker.

  45. Corkscrew

    “It will be a cracker.”

    If only you had addressed that comment to Jack – us oldies would have given you a pencil!

  46. @OldNat

    You don’t have to be as ancient as yourself to remember Crackerjack. Although most folk born in the 90s and beyond will probably think it a misprint for CrackCarJack the distant cousin to GTA5.

  47. @ Lefty,

    By the way, where are you getting the 1950s-60s poll numbers from?

  48. Maybe it’s just me, but the highlight of this month on UKPR has to be Valerie getting piped to “First” post, and declaring her displeasure with the usurper of her rightful place atop the list with the inimitable solitary word: “Oi!”

  49. OLDNAT

    What can I say…..well done?

  50. It is still far too early to say what will be the impact of economic improvement on VIs. If we grow at 2.5% per cent or so for the next year, and there is at least some improvement in average wages (as seems likely), I would be amazed if the polls have not changed significantly by this time next year. Maybe not enough for the Tory majority, but I would still bet on biggest party. Also – what do we know about the solidity of the UKIP vote?

1 2 3 4