The second of Populus’s two twice-weekly polls is out this morning and has topline figures of CON 35%, LAB 35%, LDEM 9%, UKIP 13%. The Conservatives and Labour are neck and neck on 35% a piece. Tabs are here.

The last time we saw a poll without a Labour lead was MORI’s October poll last year. That one didn’t herald a great crossover, it was just a blip. You probably shouldn’t get excited about this one yet either – it could be a further narrowing of the polls, or could just be normal variation within the margin of error. Populus tend to show some of the smaller Labour leads anyway, probably as a result of their weighting scheme (Populus weight by party ID, in a similar way to YouGov, but weight Labour to a lower level of identification).

Meanwhile the daily YouGov poll for the Sun had topline figures of CON 32%, LAB 38%, LDEM 8%, UKIP 14% (tabs here). As ever, look at the polls as a whole, don’t make the error of looking more at the ones that give more unusual or exciting results.

On unrelated matters, nominations for the European elections closed yesterday and candidate lists were published, so I’ve updated the election guide part of the site with the full candidates: Scotland, North East, North West, Yorkshire, West Midlands, East Midlands, Wales, East of England, South West, South East and London.


269 Responses to “Populus – CON 35, LAB 35, LDEM 9, UKIP 13”

1 2 3 6
  1. no crossover yet. populus clearly is the likeliest polling company to show this.

    The lie of the land is still with labour, however. Assuming a 2-3 point labour lead, the tories still need a lead of 3%+ to be the largest party.

    possible, but difficult in little more than a year.

  2. is there any conclusion one can draw from the frequency of polls outside the normal range?

    crap polling maybe, or it might speak to something in the circumstances making normally representative samples not representative.

  3. As I mentioned on the previous thread concerning this Populus poll, Scottish figures (as is so often the case) quite different from rUK. Labour lead over SNP 7 % with Tories a close third. MoE? Small sample?

    Wait and see……… but not good news for SNP

  4. This is a truly sensation poll which will send shock-waves through the political establishment.

  5. @John B

    Quoting unadjusted cross-breaks for Scotland? Bad form, old boy, bad form.

  6. @ John: Sample size for the Scottish crossbreak is 185, so MoE is ±7.2. So I wouldn’t bother…

  7. Off topic, but did someone based in Spain try to view my site yesterday? Whoever it was got blocked, and that was a shame, so I’d like to rectify it. :)

  8. Some talk of the Greens on the last thread. Some time ago (1989? – can’t recall) Greens got 15% in the EU poll, and it was hailed as a breakthrough. Of course, it wasn’t, and all the excitement was for nothing.

    With UKIP this time, I do feel it is qualitatively different. The gestation period has been long and steady, there has been a far wider spread of results in by elections and local elections, and the media coverage is much more visible in advance.

    A couple of days ago Farage was in my region, and was the lead story on both main channels local news, top story in the regional press, and achieved a convincing level of other coverage in a traditionally unfriendly region. [That is, unfriendly to UKIP – we’re really canny folk up here in the main, if you can see what I’m saying].

    There are still big questions over how much of this UKIP support will carry through to 2015, but my suspicion is that a much higher level of support for UKIP than people thought possible in 2010 have developed, and solidified in the intervening years.

    The GE might still see them back on 5 or 6% nationally, but that extra coupe or three percent will have a big impact. I feel it could actually be a fair bit higher than this. The signs are that their vote will be highly targeted as well in specific constituencies, so I do feel that seat by seat results are going to provide a significantly bumpy ride for forecasters come election night.

  9. Does Statty have Spanish fans? Interested in UK polling? Who knew?

    ….

    maybe it’s the Catalan thing… interested in the outcome of Scots Independence!!!…

  10. @john b

    replied to you on previous thread. Just thought I’d add that after much scrolling, I did eventually find an Italian uni in the league table, down in 221st place. Must be all those “dottores” you were on about…

  11. @Carfrew

    No idea, but the referrer was a page on this site, where I had posted ‘charts updated’, so it’s probably legit enough.

    I was guessing that one of you rich politicos was getting some Easter sun.

  12. @ Anthony: I’m not sure the house effect is *that* big… I reran your analysis from last summer and I get an average difference vs Yougov of just under half a point:

    https://www.dropbox.com/s/o3zqwr379mfyx45/House%20Effects.png

    Sorry for the messy chart. Of course, my numbers could be wrong, but I got pretty close to your numbers from last summer for the same period.

    Two things to note: The pollsters seem to be closer to one another in terms of Labour leads, admittedly Angus Reid aren’t in this sample, but the others still seem more tightly bunched.

    Secondly, the pollsters seem even more polarised in terms of UKIP! While the difference between the highest and lowest is still around 7.5 points, the new online pollsters have tended to converge on the top of the chart (TMS BMRB only seem to have done two polls during the period).

    The second point begs the question: Why the online vs phone difference? Shy UKIP voters or stuffed panels? Personally I don’t think UKIP supporters don’t seem like the shy type, but it would be interesting to see how online panelists recruited before the UKIP surge compared to those the joined since it, to see if there is any residual bias after demographics and the political weightings are taken into account…

  13. @ Crossbat

    People don’t do irony- you should know that by now :-)

  14. @STATGEEK

    “I was guessing that one of you rich politicos was getting some Easter sun.”

    ——–

    Fat chance. I’m getting british rain…

  15. @Nigel.

    I wouldn’t actually say “crap polling maybe”
    Polls are as you know “just are a very very small sample of voters”. What is clear though is that the Labour lead has fallen quite dramatically over the last 4-6 weeks. We have a few pollsters that in the last 4-6 weeks have been showing the parties level or 1-2 points apart ie. Opinium, Populus, Survation even Yougov have shown very small leads of 1-2%.(Have a look at Anthony’s Voting Intention Opposite) So with a +/- 3 % margin of error we could already be in a crossover position or a small Tory lead.

    Its best to wait and see if this trend continues over the coming months before drawing any conclusions. I might be wrong (not usually).

  16. @ Carfew – Not sure if that was a serious comment about the Catalan thing, but to me it’s not far fetched at all… I got a lot of questions about Scotland while I was in Barcelona recently!

  17. @Shevvi

    “People don’t do irony- you should know that by now :-)”

    Understood. I gather ironing is the preferred pastime for some posters.

    :-)

  18. @Number Cruncher

    As Freud said: “There are no jokes”…

    (well, someone told me Freud said it, he may have been joking)…

    But yeah, the Scots peeps take an interest in the Catalan thing, so why not the other way around. Statty needs to be on his best behaviour ‘cos he is representin’ us now…

  19. Time to update the UKPR ?

  20. @CROSSBAT11

    “Understood. I gather ironing is the preferred pastime for some posters.”

    ——–

    Nope, steamer all the way for me. Did find myself curiously drawn to the allotment section in Sainsbury’s yesterday though. They have allotment sections now apparently… which are not a million miles from the wine section…

  21. @Alec

    Think you’re right.

    I’m beginning to wonder if Ed M is ignoring UKIP tactically, on the principle that UKIP can do more damage to the Tories than he can, so why should he stop them doing it.

    I’m reminded of the German collusion with Lenin, aimed at undermining Russia, and the Anglo-French collusion with Arab nationalism, aimed at undermining the Ottomans, during the first world war. Tactics with far reaching and disastrous consequences!

  22. Populus clearly have a problem. As I think Phil Haines first pointed out, the figures for recalled 2010 vote are consistently out of skew with what the actual results of that election were.

    This becomes particularly clear if you look at way the recalled vote for Con, Lab and L/D is split between them. In this poll the figures are

    Con 43.5/43.3 (40.9)

    Lab 29.1/28.7 (32.9)

    L/D 27.4/27.8 (26.1)

    The first two figures being before and after weighting and the actual 2010 percentage in brackets

    So the sample contains ‘too many’ people who say they voted Con or L/D and too few people who say they voted Lab. And Populus’s weighting does little to correct this – indeed may make it worse.

    This isn’t a one-off effect caused by a rogue sample. If you look at the combined figures for March (with a sample of 16,424) you can see the same effect:

    Con 42.6/42.9 (40.9)

    Lab 29.9/29.2 (32.9)

    L/D 27.5/27.8 (26.1)

    I suspect the rationale behind this is that some people are suffering from false recall being less likely to say they voted for now-unpopular Parties such as Conservatives and Lib Dems and more likely to claim they supported Labour and UKIP whose ratings are now higher. But in that case you’d expect the unweighted figures to be more pro-Labour[1] while they are still under the vote they actually got. And that’s before Populus then re-weight to reduce the influence of 2010 Labour voters even more.

    I think the problem Populus have is that they are still expecting their respondents to behave as if they were in a telephone poll. But with a telephone poll there may be a greater impulse to please the real life interviewer by giving a response thought more ‘acceptable’. That pressure is not just less online, but panel members may feel that that the system already ‘knows’ how they voted from previous answers and so be more likely to not change their reply[3]. By applying phone-style adjustments[4] to online polls pollsters may be reinforcing a bias rather than correcting for one.

    [1] This is true to some extent of UKIP. The March figures show as 4.3% of the Four Part Split rather than the 3.4% in practice (obviously you need a full month’s sample to get a decent number of 2010 UKIP voters). But Populus re-weights them down to 2.8%, going the other way.

    [2] We saw this with their first attempts at online polling starting last July when they got UKIP’s vote ridiculously low. They seemed to be trying to fit UKIP’s figures to the levels they were finding in telephone polling when they stopped that the previous January. Unfortunately the UKIP vote had increased since, boosted by their performance at Eastleigh and in the county elections, so Populus’s figures were clearly out of line with everyone else.

    [3] There may be other reasons for greater accuracy as well. We know panel members are more likely to be politically aware than average and therefore more likely to remember. The very process of having given the information may also fix it better in people’s minds.

    [4] Phone polls also traditionally tended to find more Labour voters, while the same may not be true of the members of online panels – indeed the bias may be in other ways.

  23. @Carfrew

    Actually Freud wrote a book called “Jokes and their Relation to the Unconscious”, which is a rather interesting read.

    Anyway, we know he had a sense of humour. “Sometimes a cigar is just a cigar…”

  24. Carfrew
    I’ve left you a little message on tpt.

    Wayne
    Should we call you ‘Not usually ‘?

  25. @EWEN LIGHTFOOT

    “Carfrew I’ve left you a little message on tpt.”

    ———

    Yes, you lowered the tone by mentioning the people who punt from the wrong end. Have recovered and left you a reply tho’…

  26. @POSTAGEINCLUDED

    Actually Freud wrote a book called “Jokes and their Relation to the Unconscious”, which is a rather interesting read.

    Anyway, we know he had a sense of humour. “Sometimes a cigar is just a cigar…”

    ———

    Thanks for the heads up, Postage, I shall add it to my Ukpr booklist for future perusal. But I thought he was being serious about the cigar thing. I mean, sometimes it is just a cigar, isn’t it? Or have I been doing it wrong?

    (…Freud prolly never met Clinton though…)

  27. @Carfrew

    I’ve always imagined him sucking lasciviously on the said Havana before delivering the line in the manner of Groucho. Possibly in Yiddish, which is always funnier.

  28. Me, thirty years ago teaching A level Psychology to my first class, and trying to get across the plausible , yet un-scientific nature of Freud’s model of the personality…
    Voice from the back of the class, “So, Freud was just a Bullsh*tter was he Sir?”.
    Happy Days !

  29. @Wayne

    “So with a +/- 3 % margin of error we could already be in a crossover position or a small Tory lead.”

    If you’re right, and I tend to think you’re not, wouldn’t we have seen some Tory leads by now? If the Tories have a small lead now, as you suggest they might have, then the latest YouGov showing Labour 6 points ahead is an absolute shocker of a rogue poll isn’t it, as would be the one showing a 5% Labour lead the day before?

    You mention other pollsters, but while some have shown a narrowing Labour lead, particularly Populus, none have showed the Tories ahead. Populus had Labour leading by 3% a few days ago.

    You also referred to Opinium a pollster supporting your case. Their latest poll showed Labour with a 6% lead.

    [Snip]

  30. Here we see some small evidence that a Conservative OM can’t be ruled out. I’m thinking we will see more polls like this over the summer. Things did drift a little for Labour last summer, only for Milliband to strike a nerve at conference and regain much lost momentum. It seems reasonable that a similar pattern will emerge this year.

  31. “So with a +/- 3 % margin of error we could already be in a crossover position or a small Tory lead.”

    Clearly not if you look at current polling averages of a 3-4% Labour lead.

  32. Lol, Populus. But it gives us the pleasure of some rare sightings here at UKPR, like Sine Nomine, so long may their mystifying weighing scheme continue!

    I went back and looked at their European election polling, and I noticed they were one of the few pollsters who have the Conservatives ahead of Ukip. (Not entirely surprising, considering how they upweigh Tory voters.) So we shall have an empirical test of their methodology soon enough.

  33. @Alec 1.44

    We’re canny folk up here…..

    Your mean ‘down here’, surely!

  34. Incidentally, Labour are down a point or so pre-Budget with YouGov. Since the monthly/bimonthly polls are too infrequent to reveal such a small shift on this timescale and Populus is Populus, I think we have no choice but to place some trust in the one tracker which seems reliable, and that’s what YouGov is telling us.

    I don’t fully understand what prompted the shift, but it has happened.

  35. @KeithP

    No overall Tory majority possible on a poll lead of 3%. They need at least 5, if not 6, depending on what other parties do, because of the lack of redistribution.

    @Postageincluded 1.05 p.m.

    I apologise: I thought I was reading the adjusted figures. The unadjusted ones looked even better for Labour. However, as others point out, Populus is not entirely everyone’s favourite polling outfit.

  36. @Wayne

    “What is clear though is that the Labour lead has fallen quite dramatically over the last 4-6 weeks..”

    Actually no, that isn’t clear at all, and just a simple average is enough to show you that – ie polls for the last 3 weeks have average lead of 4.1; for the 3 weeks before that the average lead was 3.6. Not sure I’d call 0.5% a “dramatic” or even a “significant rise, but a rise it is, not a fall, dramatic or otherwise.

    The only way to avoid just seeing what you want to see is to do some maths.

  37. @” However, as others point out, Populus is not entirely everyone’s favourite polling outfit.”

    lol :-) :-) :-)

  38. does anyone have the monthly averages for the last 6 months for yougov handy?

    If not, no worries I’ll just get my pen and paper out :)

    Al

  39. @Carfrew 1.55

    Yes of course, Oxford is going to turn out very high quality students and I’ve no doubt that many other universities around the world would love to be able to demand such high entrance standards.

    My point was rather that a normal UK first degree will not usually cover the ground covered, for example, by a German first degree, simply because of the timescales involved. Oxford sacrifices breadth for depth. This produces people who know virtually everything there is to know about one aspect of the chosen subject – and I think there may be more ‘world authorities’ in Oxford than in any place in the world; extraordinary place.

    But an ordinary student who studies for five years in an ordinary university in Germany or Italy (not many Italians make it through the entire course, but that’s another story) will have a broader understanding of the wider subject that one doing a three year course in an ordinary UK university. Normal UK students only make it to German degree standards when they take a second, Masters, degree. At least, that’s true in my own field….. and my impression is that it is accepted as a general rule.

    But, as ever, I’m happy to bow to superior knowledge….

  40. This banter about whether we have CrossOver or not – sounds like something from a science-fiction film – seems to have a lot less vitriol about it that the discussion on whether a certain other set of polling figures show the gap closing or not.

    Just thought I’d mention it……..

  41. COLIN

    LOL indeed!

  42. How are you TOH?

    Weather is a bit grim here today-much colder.

    Good Retail numbers on ONS today.

  43. @Carfrew
    Yours of somewhere around 2.30 p.m.

    “Lol, Ewen, John b was putting the Italians above us, I was just evening things up. France’s first showing is in at number 65, which is quite good, all things considered. Dunno if you get called doctor there though…”

    No I wasn’t putting the Italians above the UK. All I was pointing out – and ought to have been much clearer about it – was that if an Italian finally does gain a degree from one of its better public scholastic institutions (which, by the way, have all suffered horrendously from cuts in government spending over the past ten years) it is after a longer period of study, covering a broader range of subjects than is common amongst UK courses.

    ON the other hand, you may choose to take what I originally wrote as a feather in the cap for UK universities – that a simple bachelor such as myself might be regarded as of a standard accredited with the title ‘dottore’ in another country.

    Quite honestly, I’ve forgotten now how all this started……..

    Anyway, the weather is pretty bad here as well though the fog seems to be lifting now…..

  44. @RM

    Thank you for that analysis of the Populus polls, it is the kind of contribution that makes this site so interesting. I am at a loss to understand anyone being excited by this poll when they are consistently out of step with all other polls. I say this as a member of their panel who has not been polled for over a month so presume they have no problems with sample sizes.

  45. The fact that without a crossover, they still might not win an OM suggests to me we should probably look at having a fair voting system. It’s 2014….

  46. @ RogerMexico

    Can you explain the 3 %ages you noted for each party again. I didn’t follow (especially the one in brackets)!

    @ Colin and TOH

    What are you two sniggering about? :-)

  47. Colin

    Thanks for asking, feel fine, get my results on 29th. Hope things are well with you. I look in most days but not posting much at the moment, just too busy. Wet day today so we spent a pleasant couple of hours at the National Gallery.

    Most of the recent economic numbers have been good and I suspect the effect is slowly filtering through to the voters. We shall see.

  48. @Zack

    Sorry – don’t understand. Surely you mean “even with a crossover they sill might not win an OM” (they being the Cons?)

    But surely having a fair voting system would take all the fun away.

    More seriously, though, what would count as more fair? Proportional representation? But then what happens to the constituency link with the MP, such an important part of our UK system? And it would make OM government almost impossible to achieve.

    On the other hand…….

  49. @TOH

    All the best for 29th.

    As for economic stats, if where I live is anything to go by it will take a long time for any effects to come through – certainly no sign of anything yet.

  50. “UKIP have clashed with the Tories over their use of an Irish-born actor on an election poster about British workers being hit by foreign labour.”
    I suppose the Irish don’t count as real Europeans

1 2 3 6