Tonight’s YouGov poll has topline figures of CON 40%, LAB 39%, LDEM 9% – a Tory lead but still in line with the recent average of a small Labour lead of a point or so.

Opinium have also released new figures (though the actual fieldwork was conducted last week). Their topline figures with changes from early February are CON 35%(-1), LAB 39%(+3), LDEM 10%, Others 19% (the very high others score is typical for the company, for reasons which are not clear).

We are also still awaiting Ipsos MORI’s monthly poll for Reuters which I thought would have turned up by now.

40 Responses to “New YouGov and Opinium polls”

  1. Is there a theory that explains the tendency of Yougov to show a Tory lead of 1 every two to three days?

    Some peculiarity in the polling?

    Political simple harmonic motion?

  2. I suspect we are in a slightly simlar position to last month in terms of polling – with some polls showing 1 point lead/roughly neck and neck figures for Lab compared with the Tories and another set of polls showing 2 – 3 leads roughly (TNS BRMB and Angus Reid come to mind regarding last month). I suspect the 1 point Tory lead confirms it’s still roughly neck and neck with a 1 point Lab lead.

  3. I note that Opinium say that ” The sample is scientifically defined”.

    I doubt that very much. It has no doubt been statistically defined, but to claim that sampling populations is “scientific” suggests a lack of understanding of what science is.

  4. So, it’s time for the obligatory YouGov 1% Tory lead that seems to turn up every three or four days! I think I’m detecting the pattern now; if the day starts with a T and ends with a Y, then it’s a TorY poll lead!

    The Opinium poll is interesting, though. They had the Tories 2% ahead at the end of last month and the parties all square only two weeks ago so a 4% Labour lead is quite a turnaround inside four weeks.

  5. “Is there a theory that explains the tendency of Yougov to show a Tory lead of 1 every two to three days?”

    Neck and neck lead swapping polldrums as stated for a few weeks now.

    Both parties (on YG) seem unable to break through 40 for more than a day or two (i.e. they can’t outside of variance or sample error).

    Whether they have convinced their right flank of golden victory in Europe (quite funny when you look back on al that froth and listen to the emergency motion today); or seem convinced that each ‘terrible story for the government’ is going to be the killer blow that send Ed to mid/upper 40’s…..until the polls illustrate that it is not!

    We have an electorate that is ostensibly split down the middle as some of us first asserted back in late 2010. Nothing has basically changed since then.

    IMHO I doubt it will by the next election (sometime between late 2013 and May 2015).

    As each day and week and month passes it looks more and more likely- new or old boundaries- IMHO that we are headed towards the second hung parliament in a row.

    Just with a LOT less Lib Dems….

  6. The problem is, YouGov spoils us.

    YouGov can go from a 2% Tory lead, to level pegging, to a 4% Labour lead and back again in the space of a week.

    It makes it hard to get excited about Opinium’s movements, which simply feel like YouGov but painfully slow….

    Everything still just points to a 1 or 2 point Labour lead, with statistical wobble around.

  7. @Crossbat.

    I suspect that’s what you’ll notice with those pollsters who had results at the top of MOE at that time (which was 1-5 Tory leads). With other pollsters such as ComRes, and Populus who were more closer to the mark last month, their polling will produce quite small changes, by contrast. So I suspect this is Opinium slightly falling into line with the current trend of a Lab lead, though albeit currently at the top of MOE now which is 4/5 point Lab leads.

  8. Good Evening. Just in from a long day of teaching.

    Did anyone see the tough interview with Andrew Lansley and Nick Robinson?

  9. Tonight’s details:

    Con 40
    Lab 39
    Lib Dem 9
    UKIP 5
    SNP / PCY 3
    Green 2
    BNP 1
    Respect 0
    Other 1

    Approval 30 – 54 = -24

    Non-voters 24%

    Tables are here:

    Non-voters still high-ish by usual standards, but nothing noteworthy. Only change on second lot of policy trackers is the expected rise in Health. AS Boo Boo says just a reflection of an underlying 1 point (maybe plus a little) lead.

    (Boo Boo margin or error on last nights figures was around +/- 2, varying with policy area – because it’s smaller if the percentage being estimated is lower. So 3 would have been significant).

    I wonder if the MORI poll is being re-positioned so it comes out earlier in the month and doesn’t get lost with all the others. Mind you they missed out December, January and February last winter, so maybe it just a reduced hibernation (global warming affects everything).

  10. @Neil A – “… hard to get excited about Opinium’s movements”

    However, Con on 35% looks fairly consistent with Opinium polling over the last year.

    Also of interest – beyond the movement – is determing the absolute level of VI. Companies which report less frequently still show fairly consistent trends… but they find the VI to be within a noticeably different range.

  11. Youguv swinging around as usual…I notice that with other polls,the Labour seems to have increasing leads…ICM-Labour ahead by 1,Populus- by 2,ComRes-by 3,Opinium-by 4.
    I half-expect Ipsos-Mori to show a 5 point Labour lead

  12. @Roger Mexico

    I’m still not essentially getting your argument which basically seems to because the averaging was +/- 2 3 is therefore signifcant. It’s only about a point’s difference.

  13. @ChrisLane1945

    I saw the tail end of the Lansley/Robinson interview on the BBC 10 o’clock news tonight and you know a Tory Minister is in desperate trouble when the likes of Robinson go for the jugular! I haven’t seen the hirsute-less one so discombobulated since it looked as if Brown, Mandelson and Campbell were going to stitch up a rainbow alliance coalition in May 2010. What do you think his problem was? Probably Lansley’s failure to communicate the reforms effectively. There’s nothing a Tory likes less than a fellow Tory letting the side down!! lol

  14. why is your poll of polls and latest polls running a week behind the time?

  15. @Smukesh. I’d again I’d state was I stated to Crossbat on Opinium’s poll movements. I’d say the same for ICM as well (although their Lab is within current trend). I suspect we are seeing from polling companies such as ComRes, Populus, Opnimum etc monthly snapshots of the swing abouts and round abouts YouGov go through every week which roughly averages 1 point Lab lead when taken in with the 1 point Tory leads we see every so often as well.

    It’s simlar to last month in which we had group of polls which had the more higher Lab leads (TNS BRMB and Angus Reid) to the pollsters which had the higher Tory leads (ICM and Opinium) and other pollsters showed roughly neck and neck. All were a collective snapshot of YouGov’s swing abouts and round abouts for a 1 point Tory lead. Now, we have largely a simlar picture. A group of polls which shows roughly 1 point Lab leads, the roughly neck and position we see, the higher Lab leads of 2-4 we sometimes see, and the odd 1 point Tory lead shown by both ComRes and YouGov and different points indicating the still quite close position. All and all, in a month nothing much in polling has actually changed – we still pretty much neck and neck.

  16. CROSSBAT11
    I have often wondered about Nick Robinson.

    On another matter: the BBC reported on National news the violent death of a Year 11 boy in one of our schools here during school time. May he rest in Peace.

  17. In his interview with Nick Robinson, Andrew Lansley has said that he has changed his mind about the benefits of competition in the NHS. I wonder if he will now be taken out and shot out of the Cabinet.

  18. Conservatives should have just let privatisation and competion creep up naturally from the starting point Labour introduced, which was circa 10%. Instead they are going to get blamed for it all if it goes wrong.


  19. BOO BOO
    All and all, in a month nothing much in polling has actually changed – we still pretty much neck and neck

    Yet 4 of the 6 polls show Labour increasing voteshare and Conservatives down or level.AW thinks that Labour has a small lead.
    If next months` polls show a Tory lead,then I`ll agree with you that they are neck and neck.Till then I`ll believe that the net movement in polls this month has been towards Labour,possibly from Tories,possibly from other parties.

  20. Why does the vote share in the Opinium Poll add up to 103%?

  21. @Smukesh

    Again. I addressed all your points in my previous post that you mention. I’d say a swap between a 1 point Tory lead to a 1 point Labour lead is not much of a change at all – and that a 1 point gap pretty much indicates it’s neck and neck. Again, on your point regarding an increase in Lab’s vote share in these polls – I addressed that as well. Indeed, I noted several weeks back I believed the Tories had lost a point, a Lab had gained a point which gave the impression for the underlying movement you and others’ are calling on. Most polls certaintly confirm my 1 point drop in Tory lead call, pretty small and insignifcant all in all.

    All the other polls which shown an increase in Lab’s vote share (averaging roughly 2% which would align with YouGov’s 1.5 or so Lab lead averaging) and the overall movement, again depends on where they were a month ago. Although last year, I critcised daily polling, I’ve now come to support it because it gives a much easier way of picking out the outliers, and the MOEs and spotting the underlying picture over sample variation. Polls last month, like this month show varying snapshots of the ups an downs YouGov go through, the MOEs caused by sample variation to the rough averaging.

    ICM themselves are roughly calling the averaging. The big movement from them is simply due to the fact their results in January were at the top of of MOE at the time, rather than the rough averaging of 1 point Tory lead – of which the movements would have looked minimal like they did in the polls like Populus which showed Tories down 1, Lab up 2. Opinium is simlar to ICM, in this respect except they’ve simply gone from one top of MOE to another! As for net movement to Lab – that’s likely to be from the LDs who have exprienced a slight drop in their voteshare as well.

  22. On the Opinium page Others total 17%.

    Still adds up to 101% though.

  23. under the new system it could be, assuming polling like this continues, that we get the two main parties with around 280-285 seats each and 10-15 LD’s: that means a two-party coalition for government isn’t going to have quite enough seats. Perhaps we’ll see 3 parties next time?

    Something interesting going on in Ireland, apparently they want to hold a referendum on the new fiscal compact (mini treaty, whatever you like to call it), apparently Germany wants to push the thing through without much further discussion. We know what happened last time such referenda occured in Ireland.

  24. KeithP

    They kept asking until Ireland answered the correct way?

    If Ireland does say no, it won’t scupper anything the mini treaty can have people opting out if the wish. It will become one of the requirements if Ireland needs a new bailout in future though.

    Pretty much would be the end of their 12% corporation tax I think.

  25. @NEIL A

    “YouGov can go from a 2% Tory lead, to level pegging, to a 4% Labour lead and back again in the space of a week.

    Everything still just points to a 1 or 2 point Labour lead, with statistical wobble around.”

    Agreed. My MAD (60-poll) figures for the UK have it at:

    Con 39.0
    Lab 40.1
    Lib 9.2
    UKIP 4.8
    Green 2.0
    Other 5.2

    A 2% shift would give the figures you mention.

  26. I’ve been watching the polls on here without fail every day for the last few years. The poll of polls is not updated until there is a Tory lead in a poll or two. It can go for a couple of weeks at a time or more. It would be good if the poll of polls was updated more regularly. How about once a week? The last time before the most recent update was 8/2/12 when there was a +2 Tory outlier.
    Anthony Wells – please explain how the poll of polls is calculated and when. Why is it not updated on a weekly basis?


    That seems a tad demanding!

    It’s Anthony’s site, and it maybe that you have never had a new baby in your house?

    I’m sure that if you want to chuck a few thousand quid to Anthony every week, he’d be happy to comply, and even answer a cheeky “please explain” comment! :-)

  28. Boo Boo (10:34)

    Sorry I didn’t make myself clear. ‘Significant’ here is a bit of statistical jargon. It means that it’s outside the limits which you would expect 19 out of 20 samples to fall in. If you were looking to estimate the support for something (which had a true value of 25% in the population as a whole) by taking samples from the population of 1750, then 19 times out of 20 you would get a value between 23 and 27 (25 +/- 2). Anything outside that range would be declared significant at the 95% level. Here the ‘2’ (actually 2.03) which is the margin of error has been worked out using a calculator program:

    So a drop (or rise) of 3 points or more in saying the Conservatives are best on a particular topic in a poll with this sample size and with the ‘true’ value of 25% would be significant in that sense, particularly when it happens in the majority of topics and in a sample where the Conservative VI is unaltered. As I said how long the movement lasts and how significant it is in political terms is another matter.

  29. JSFL and Billy Bob

    The topline figure has to be a typo, probably in calculating the Others, but the old rounding problem means you can get your percentages to add to 100 + (the number of categories/2). Suppose you had:

    Con 34.5 rounds to 35
    Lab 38.5 rounds to 39
    L/D 9.5 rounds to 10
    Oth 17.5 rounds to 18

    So 100% rounds to 102.


    Con 34.4 rounds to 34
    Lab 38.4 rounds to 38
    L/D 9.4 rounds to 9
    Oth 17.8 rounds to 18

    So 100% rounds to 99.

    Opinium don’t seem to publish tables (at least I couldn’t find them) so I can’t say this is what happened but it happens all the time with most pollsters.

  30. The Times today has teasers for a new Scottish poll. Short version: 5-year outlook not seen as very good for incomes (+4), job security (-1) or Scottish economy (-5) in case of status quo, worse in the event of independence (-14/-7/-11). Some good news in that support (+16) for the idea that independence would be good for Scotland’s standing in the world. Based on past Times coat-trailling that’ll probably be the worst they could come up with, so I’m as complacent as ever. No tables up on the IPSOS/Mori site as yet but the online article has a link so they may be soon.

  31. @KeithP

    You said “…Something interesting going on in Ireland, apparently they want to hold a referendum on the new fiscal compact (mini treaty, whatever you like to call it), apparently Germany wants to push the thing through without much further discussion. We know what happened last time such referenda occured in Ireland…”

    Oh lord, I told you earlier in the year there would be an Irish referendum (it was one of my year predictions). The treaty requires changing each country’s constitution, but the Irish constitution cannot be changed without a referendum, so an Irish referendum was inevitable. If the word on the Irish street is correct the referendum will be lost, so Ireland will be locked out of the ESM and won’t be able to get an EU bailout the next time it needs one. So it’ll have to go straight to the IMF instead. Points to note:

    * 1) Will this result in the collapse of the treaty? No, because the treaty is designed to achieve ratification when twelve countries ratify it.
    * 2) Will Ireland be asked to vote again? Not if it doesn’t want a second bailout, no (and so far it doesn’t). This treaty is (not accidentally) Ireland proof.

    Points to note
    * 1) Why is it de rigeur to call this a “fiscal compact”? It’s an agreement between nations to do something. The word for this is “treaty”. It even calls itself a treaty and has done so since January 6th.
    * 2) This is moving fast. In my predictions for 2012, I said the treaty would be signed by May/June, the Oireachtas would ratify in October, the Irish referendum would be after that and fail to be passed, Ireland would fail to adopt. But the treaty will be signed by the end of this week (!), the Irish referendum will be before May (!!), and the treaty will be fully ratified and in place by about June (?). By EU standards, this is warp speed.
    * 3) Why twelve countries? Under enhanced cooperation, it only needs to be nine/ten (one-third of EU27) Why the extra two?

    Bit I got wrong.
    I thought it’d be called the Treaty of Copenhagen, Brussels or Nicosia, depending on when/where it got signed. But no, it rejoices in the name of the “Treaty on stability, co-ordination and governance in the economic and monetary union” (TSCG for short). Fair trips off the tongue, doesn’t it?

    Regards, Martyn

  32. @ Martyn

    “Treaty on stability, co-ordination and governance in the economic and monetary union” (TSCG for short). Fair trips off the tongue, doesn’t it?
    They want everybody (except Ireland?) to sign it, then never speak of it again. :-)

  33. It is odd how “Europe made Ireland vote till they got the answer they wanted” meme has taken root here. This simply isn’t true.

    Ireland have a lot lot more referenda than the UK has. (Ten since 2000) And a vote in one is not considered “settling the matter for a generation” as it is here. The same subjects have come up multiple times, and this is even recognised by recent practice of reserving the numbers of failed amendments so they can be reused for a new referenda on the same or substantially equivalent subject. It occurs quicker with the European treaty accession referenda, because the cycle of ‘reject, renegotiate, re-propose’ is quicker after a no vote on a treaty. If there was ever a treaty where the second vote was no, it would be very unlikely to go back for a third!

    LAB 41 (+3+, C 35 (-3), LD 12.
    I think it is the first poll in February with a 6 point LAB lead.
    I have been very closely watching IPSOS polls for France, because in the end of February 2007 it was the institute that most accurately predicted the outcome of Presidential Election of that year 53 days before 1st round and 67 days before the runoff. The IPSOS VI poll of that day had Sarkozy 32, Royal 25, Bayrou 18 and for the runoff Sarkozy 53.5 and Royal 46.5. Actual results were Sarkozy 31.1, Royal 25.9, Bayrou 18.5 and at the runoff Sarkozy 53.1 and Royal 46.9. So what is IPSOS VI poll for this year at exactly the same moment of the campaign? Hollande 31.5, Sarkozy 27, LePen 16, Bayrou 11.5, Melenchon (rad. left) 9.5. Runoff: Hollande 58, Sarkozy 42. So, if the past teaches us well, Sarkozy is toasted. I remember back then waiting with anguish for the slightest fluctuation in VI, and indeed there were some polls during February-March, mainly by institute CSA, having Royal at 49-50, giving us socialists some hope (which, of course in hindsight was a vain one). But now Sarkozysts have not even that, there is not a single poll having Sarkozy higher than 27 at first round and 45 at the runoff, which is in itself astonishing for an incumbent president.

  35. VIRGIL
    Good Morning.
    The French Left has a credible candidate. Peut etre the Labour Left will one day get a credible leader; someone that can ‘speak to England’.

    Normally it takes three terms of tory government for this to happen. This time around the fib dems are shoring up the tories and voting for all their measures, forming a centre right block which will be formidable, Credo/IMO in 2015.

    We have had a critique from Lord Hattersley on the Left here. I am trying to find out if he is related to the Hattersley who was a senior figures in the seven debacles from 1970-1992. Do you know if this is the case?
    (Also being a grammar school boy himself, helping his hero to destroy every f*****g in England)

    Best of luck to Francois.

  36. @Chrislane1945
    If you are referring to Roy Hattersley who recently published in the Guardian an article about Ed and David Miliband, defending the former’s positions, yes he is THE Lord Hattersley, who played a major role in the Labour Party in past decades.
    On the problem of strong leadership in general, I think that , although having a popular leader who strikes the right cord with the public certainly helps, UK and France differ, because in France the presidential election is the more important one, and the subsequent GE usually confirms the results of PE ,giving OM to the presidential party. In UK party policies matter more, and in fact you have 650 parallel elections, where a slight swing in the marginals can change the final result. Technically, PM in England, as in most European countries, is not “elected” by the voters, but by MPs, whereas in France he is appointed by the President, who is directly elected by the people, so the personality and popularity of a leader are very important, the socialists had won all intermediary, local and regional elections between 2003 and 2006, yet they failed to win in PE 2007 because Sarkozy was perceived as stronger and firmer leader than Royal (whom I personally admired, in 2007 it was the first time I wept over a defeat, and yet I have experienced many of them, as I vote in 2 countries since 1981!)


    @”The French Left has a credible candidate. Peut etre the Labour Left will one day get a credible leader; someone that can ‘speak to England’.”

    Mr Hollande-in London to campaign for the votes of 300,000 expat French City workers, proposes top rate income tax of 75%.

    FH meets with EM whilst in London.

    Oh yes-please let us have a “credible” Labour leader with policies like that -Dieu je t’en prie :-)

  38. @Roger Mexico

    The Opinium page gives:

    UKIP 6, Green 5, SNP 4, BNP 1, Plaid Cymru 1, Other 0, already rounded to 17%… they then quote 19% for Others in the headline figure (so a rounded total of 101%).

    I would quite like to see the decimal places, but I suppose it detracts from readability and neatness.

  39. On the phone hacking etc the Independent has a very succinct guide to how bad things might be about to get for NI, the police…and Cameron?

    Personally I think Cameron should just express disgust at NI, disown his former pals and make some expression of regret and (barring some new revelations) it won’t make too great a dent. But (with any luck) NI are going down and we might see some more very high profile Police casualties.

  40. “The growth of competition in the NHS would not be significantly effected by the Government’s reforms , according to Catherine Davies, director of Co-operation and Competition Commission, set up by Labour in 2009 to monitor letting of contracts in NHS.

    Politicians have created “a massive amount of misinformed debate” and have failed to notice that competition law already applies within NHS .
    We have essentially been applying the same principles of competition law and competition economics for the last three years. There isn’t a suggestion anywhere in the system that suddenly everything has to be thrown open to procurement, but at the same time commissioners need to make sure they can justify doing things in the way they’ve decided.” she said.

    Under the Labour Government, patients were gradually given more choice over where they would be treated for elective operations and other non-emergency care. This had improved services Ms. Davies said, adding that the gradual expansion of choice would continue ”

    The Times.