Tonight’s YouGov poll for the Sun has topline figures of CON 40%, LAB 39%, LDEM 9%, a one point Conservative lead. This follows yesterday’s poll which had a four point Labour lead. The two main parties still appear to be pretty much neck-and-neck in YouGov’s daily polling, but the fact that we’re getting slightly more Labour leads than Tory ones, and slightly bigger Labour leads than Tory ones, suggests that the underlying position is probably a small Labour lead – perhaps a point or so.

184 Responses to “YouGov/Sun – CON 40%, LAB 39%, LDEM 9%”

1 2 3 4
  1. Hmmm, so likely a 1 point Lab lead, but it’s still pretty much neck and neck considering how close the two parties are at the mo.

  2. It does look like it’s all sample variation around an underlying slim Labour lead. Last night’s sample looked favourable to Labour (but it’s not worth tempting fate by suggesting it on here :-) )

    Tonight’s looks a little more blue friendly with a very low LD to Lab transference. So as Anthony says, it’s a narrow Labour lead, with a few oscillations depending on the sample….

    Where we were really……..

  3. Yawn

  4. As Robin said yesterday, the gap seems to be widening.

  5. Any reason why we only seem to be getting some poll results.Why are the couple before this one not up?.

  6. Lab and Con roughly equal in England. Lab and SNP roughly equal in Scotland for Westminster voting.

    Due to the collapse of the LDs, Labour well up in England compared to 2010, SNP well up in Scotland compared to 2010.

    While I’d hesitate to endorse Jim Jam’s comment ……..

  7. @Sergio

    Indeed. The spread is now between +2 to Con and +5 to Lab, putting the Lab lead at about 1.5. Or we can take the mean VI for the last 10 polls, which puts Lab at +1.4. Either way, it’s the biggest lead since the ‘veto’.

  8. An interesting development on the health issue. The drop the bill petition has been crawling slowly upwards over the last few weeks but it seems it has just gone viral, putting on 10,000+/day now. The power of facebook?

    Maybe it isn’t the sort of poll Anthony approves of (cancel that maybe!) but it certainly is an indicator of the public mood. To my mind it reinforces the view from recent Yougovs that many people are only just now finding out about the issue for the first time.

  9. ”the gap seems to be widening”

    Not really. Lab have a 1 point lead – the same thing that’s been happening for about three weeks now. We are going into our fourth week (and thus nearly a month) of this. If the gap is widening, and a bigger Lab lead is developing, Lab leads should be appearing consistiently, rather than slightly more than small Tory leads.

  10. @Robin

    A five point lead consistiently, would be the biggest lead since the veto, not a 1 point lead.

  11. @Boo Boo

    Learn some stats.

  12. @Robin

    You didn’t really address my point then beyond some vague sentence (btw, I well aware of stats).

  13. @Boo Boo

    Yes, a 5 point lead would be bigger than the current 1.5 point lead. NSS.

    Being aware of stats doesn’t appear to be the same as understanding it.

    The current true lead is estimated by individual polls. We get a better estimate of the true lead by using a bigger sample size, which we achieve by pooling multiple polls. Doing that for the last 10 polls gives a Labour lead of 1.4, with a pretty small margin of error. That’s the biggest such lead since the ‘veto’.

  14. @Robin

    No, a five point lead would be the biggest Labour lead since the veto (as they were polling five point leads/four point leads before the veto). The 1.4 lead is a lead, but it is not biggest lead since the veto. You analysis of Lab’s average doesn’t really change that.

  15. @Hal

    An interesting development on the health issue. The drop the bill petition has been crawling slowly upwards over the last few weeks but it seems it has just gone viral, putting on 10,000+/day now. The power of facebook?

    And the power of UKPR! Thanks for the link. Just signed petition. Maybe it’s my age but I did find the CAPTCHA codes really hard to read.

  16. We enter another period of stasis in the polls.

    Not sure why, but public opinion seems to shift over events like the election that never was or the veto but otherwise stays remarkable immobile these days.

  17. @Boo Boo – I’ll go with Robin on this one. If you read his second post, it’s clear he isn’t talking about single polls but the ‘true lead’ as he puts it, as defined by a rolling average.

    While I haven’t checked his figures, statistically speaking I imagine he is correct, and that we currently appear to have the largest Labour lead since the veto if you strip out the normal MOE variation.

  18. @Boo Boo

    Yes, a 5 point lead would indeed be bigger than a 1.5 point lead. I’m glad your mathematical abilities stretch to telling which of two numbers is larger. It’s a shame that appears to be about your limit.

    What we see in polls is an estimate based on (pseudo)random sampling, whihc has an inbuilt error. Day to day polls will bounce up and down due to sampling error. A 5 point lead in a single poll does not mean that the true position is a 5 point lead.

    I’m talking about the more accurate estimates you get from averaging multiple polls, compared to similar measures from previous averages of multiple polls. The current estimate of the Lab lead, from such averaging, is the largest such lead since the ‘veto’.

  19. @Alec

    I did read his post, but I’m not borthered to really argue it any more (I can’t anyway as I have a job interview tomorrow at 9am).

  20. My verdict on the current polls, certainly the YouGovs? I’m going to put it into song: –

    “You put your right foot in,
    You put your right foot out,
    You put your right foot in
    And you shake it all about.
    You do the Hokey Pokey
    And you turn yourself around,
    That’s what it’s all about.”

    I’m off to bed.

  21. @Hal & Valerie – there is a fascinating lead story on the Guardian website tonight regarding the risk assessments of the 4 new English super regions in the NHS. Apparently they foresee significant risks of falling standards and risk of harm to patients arising from the changes, with many of the risk levels still significant after mitigation strategies. Presumably this is why Lansley is trying to prevent publication of the national risk register.

    While I’m not sure if risk registers are the kind of thing that will fire the electorate (we tend to spark off at things like dead babies and grannies on piss soaked trolleys) I would see this as very big news. It appears that the NHS themselves (as opposed to interest groups representing sections of staff within the service) are now saying that the changes will lead directly to falling patient standards and will not achieve the efficiency savings.

    I rather suspect Burnham’s point about Cameron putting the need to avoid political embarrassment ahead of the needs of the NHS could turn out to be telling.

  22. @Robin

    If you really have add in an analysis of my mathematical abilities in your post, then really illustrates you aren’t worth arguing with. I am well aware of polling averages, so you don’t have to explain it to me. It appears in my tiredness I got confused with before the veto to since veto, so I can confirm as someone who got A* in her maths GCSE that my maths abilities are one of high standards. It’s a shame your debating courtesy, isn’t.

  23. Boo Boo

    Best of luck.

  24. @Boo Boo

    Sorry if I upset your sensibilities, but if you want to avoid that you then when you post you really ought to (a) read the post you’re responding to and (b) write something that makes sense.

  25. Sample does look slightly (but not very) odd in a sky blue way. For example the 2010 Lib Dems voting Tory is about the highest we’ve seen. But nothing generally to discount the 1-2 point Labour lead.

    In the tracker questions, the only thing of note is that Health has jumped 7 points (to 32%) in the ‘country’ questions and perhaps more significantly 4 points in the ‘you and your family’ ones. If people are beginning to believe that the Lansley changes may affect them personally, this could be a problem for the coalition.

  26. @Boo Boo – “It’s a shame your debating courtesy, isn’t.”

    And I’ll go with you on this one. I thought @Robin could have been more easy going in his tone.

  27. @Oldnat



    I also can confirm, I usually do read posts (and write sense). Sometimes when you only get 3 hours sleep the night before, you’re eyes go blurry and you mis-read things. But, I guess I can’t be perfect like you.

  28. @Alec

    At least we agree on something! :)

  29. @Boo Boo

    If you have an interview tomorrow I’d suggest that posting on UKPR after 3 hrs sleep isn’t the best plan!

    Good luck with your interview. Try to sleep tonight :-)

  30. Moving on…
    Does anybody(Anthony) know when we might see a new regional poll for the South West? The last one was back in October.

  31. @BooBoo, @Robin, also @StatGeek en passant

    I need to point out that agglomerating several polls together to produce one big poll-of-polls is problematic, even if you go back to the raw data. The polls will be for slightly different questions and will cover slightly different periods, so you will be introducing an unquantifiable error. When you consider you are assessing something that isn’t stable in the first place, you may find the error you introduce outweighs any putative advantage.


    I note with considerable pleasure your work on previous boards, with especial reference to your plots of LabvsLib scores for various elections (I should imagine TGB & RobS would concur, since this feeds into a longrunning argument of theirs).

    However, i need to take issue with your Median Deviation concept – I’m not sure the underlying theory works (all together now: “the standard deviation of the mean is not the same as the mean standard deviation”) and when you consider YouGov issues way more polls than all the others combined, frequentist measures start to become slightly wobbly anyway. So I’m not sure it’s useful.

    Feel free to ignore me, (especially as I don’t know what the correct answer is!), but there is a part of me that went “ouch”… :-(

    Regards, Martyn

  32. @ Colin (from the previous thread)

    “Peter Tatchell on Whitney Houston-absolutely despicable .
    What the f*** has it got to do with him ?
    She has a family for gods sake.
    Exhibitionist idiot.”

    It’s not despicable. (That is unless he is simply making these claims in order to gain attention for himself and boost his own media profile….in which case it is despicable).

    @ Nick P (from the previous thread)

    “Peter Tatchell has outed Whitney as gay. Not sure why that is despicable, nor whether her having a family is relevant.

    Might well be true…might explain her problems if whe was trying to live a lie. It would be sad, if true. Cos she probably would have been happier out.”

    It might explain why she tolerated Bobby Brown for so long.

    I think it’s a little weird of Peter Tatchell to bring it up because so far as I know, he wasn’t acquainted with her.

    @ Old Nat

    “My only interest in Whitney Houston is her fantastic voice.”

    She had a beautiful voice. I don’t think I appreciated her as much in life as I appreciated Michael Jackson. I’m very sad that she died. She had a young child and she was so young. Even where she died is kinda sad. The Beverly Hilton Hotel is cheaper than crack.

    You know what I find odd? Michael Jackson, Ronni Chasen, and Whitney Houston all died on the same street and all within a few blocks of each other.

  33. Just a thought after Moody’s report today. I am beginning to think it will make little difference who wins the election in 2015 due to the financial straight jacket the country is going to be in for years to come. If Labour get in they will be unable to reverse, even if they wanted to, many if any of the changes that the coalition will have introduced by 2015. Anything with a significant price tag could well impact on the credit rating.
    I think because of the economic situation that the next election will be a very sterile affair, with little to choose between the parties, however much they try to stress their differences and polls like the above could well be the norm for years to come.

  34. @ Boo Boo

    Good luck in your interview!

  35. SoCalLiberal

    On the subject of coincidences, did you know that Mama Cass died in the very same apartment as Keith Moon?

  36. @ Alan

    “On the subject of coincidences, did you know that Mama Cass died in the very same apartment as Keith Moon?”

    I did not. But who is Keith Moon? That name sounds familiar but I can’t remember right now.

    I couldn’t find anything on Peter Tatchell in the Guardian but I did find this article on how both Michael Jackson and Whitney Houston used the same pharmacy.

    I don’t use it but I’ve gone in there to get some prescriptions filled for others. I’m not a huge fan.

  37. @ Alan

    I did not know that. Who was Keith Moon again?

  38. @MARTYN

    “I note with considerable pleasure your work on previous boards, with especial reference to your plots of LabvsLib scores for various elections”

    That’s news to me, unless by ‘baords you mean threads/posts. ‘Boards’ suggests other websites, and I haven’t registered ‘statgeek’ anywhere else. :)

    “However, i need to take issue with your Median Deviation concept – I’m not sure the underlying theory works”

    It depends what one expects to see from the use of it, and how one interprets (or does not over-interpret) the data. I’ll explain my own reasons with the usual geeky graphs. In the beginning, I wanted to fiddle around with excel and the poll data, and started taking interest after the Scottish elections in 2011. I’m really a bit of an IT geek with an interest in politics. I quickly realised that the Scottish crossbreaks are a messy business to interpret (I know…caveats etc.):

    h ttp://

    So…how to make the best of small samples? Perhaps we have to accept that few polls can be trusted beyond the day’s data, and that no poll is representative of any electorate. If thtat’s so, then we’re all wasting our time, and should take up another interest. :)

    However, I try to make as much sense from them as is possible, within the boundaries of reasonable expectations. For example, the polls tell me that unless the electorate are lying, or unless YG keep polling the same couple of thousand people, the SNP VI for Westminster would be far higher than the 19.9% they polled in 2010 (ok, I picked an easy one).

    The problem is that the outliers are often and I wanted a way to reduce the outliers. Standard Deviation is fine up to a point, but it is still inclined to be influenced by outliers. Quote from wiki on MAD:

    “For instance, the MAD is a robust statistic, being more resilient to outliers in a data set than the standard deviation. In the standard deviation, the distances from the mean are squared, so on average, large deviations are weighted more heavily, and thus outliers can heavily influence it. In the MAD, the magnitude of the distances of a small number of outliers is irrelevant.”

    I should stress (if it is not already obvious), that I am not a statistician, now do I have a statistician’s grasp of mathematics (boring! :) ). Just interested in making the most of Excel, and the polls give me ample varied challenges.

    Back to the 60-poll choice. Originally, I picked 30 polls, but I found there was a decided lack of trend in such a sample, so increased to 60. Applying the MAD, reduces the outliers weight on the numbers on a poll by poll basis, and I record said numbers in to a chart such as:

    h ttp://

    As you can see, the accumulative samples provide a hint of what may be happening, but it’s important to bear in mind that each data point represents the poll on that data and the 59 polls prior to that data, hence why I always refer to tehm as ’60-poll SD’ or similar. It’s not designed to be a snapshot of “who’s winning or losing today?”, but more “how have the parties’ fortunes changed over time?”, with the caveats as usual.

    Interestingly, just for fun today, I took the max and min values for the last 60 polls and took the average of the two, and came up with:

    h ttp://

    While it’s completely unscientific, funnily enough it isn’t too far from what we’re seeing in the rolling polls (small parties aside).

    I think the point of my efforts is to help ID the outliers where possible, and omit them, then see how things look. While I agree that:

    “The polls will be for slightly different questions and will cover slightly different periods, so you will be introducing an unquantifiable error. ”

    there’s always a little randomness in the data which reflects the changing mood of the electorate, so we kind of have to include it in the working, even if it skews what we expect to see. Things like the veto perhaps changed the VI (lest we forget that prior to that, the Con VI was in the low to mid 30s, and is now in the high 30s, so the ‘blip’ has not disappeared yet). In that sense, what started as outliers, are now trends.

  39. @ Hannah

    You might find this interesting. Ron Paul may have won the Maine Caucuses. Mitt Romney was declared the winner on Saturday by the state party by a scant 194 votes. But apparently, three of the 16 counties in the state are not being fully counted. All 3 of them won. In two of them, the local party officials transcribed the vote totals for each candidate as “0” in some towns including Kennebec, which is a fairly large city. It’d kinda be like having a UK wide caucus but not counting Birmingham or Manchester in the results. Apparently, one party official called in the results to the party too late…..therefore it doesn’t count. In another county, considered a Paul stronghold, the local county GOP chair, a Romney supporter, canceled the caucuses and announced they would be rescheduled. He did this apparently because of an impending snow storm. They’ve now been told they will not have a chance to vote.

    Now missing precincts from three counties wouldn’t neccessarily change the result. But the fact that the state GOP refuses to count the results from those counties makes me think that Paul probably won those outstanding precincts by enough to win statewide.

  40. Socal – from over here it all looks good for Obama unless a late Electable compromise candidate emerges at or prior to the Republican convenmtion. Either Obama will face a deeply flawed Romney (who is also quite right wing by hisotical standards) or a very right wing candidate with little appeal to the middle ground.
    Paul seems to be more complex with his isolationist stance no doubt having some cross appeal but still too right wing socially.

  41. When will the Euro collapse ?

    German economy shrinks by 0.2% Q4 2011.

    Greek economy down 7% year on year Q4 2011.

    EU meeting regarding Greek bailout is cancelled, as doubts are raised.

    Portugese economy shrinks 2.7% year on year Q4 2011, with 5 quarters in a row showing negative growth.

    It seems to me that the EZ cannot keep countries like Greece and Portugal, as this is totally unfair. They need to convert back to their own currencies, so they can devalue their currencies and have fiscal control. By keeping them in the single currency, this could arguably cause the whole EZ to collapse. The German and French people will not wish to keep bailing out countries.

  42. austerity is going to fail in the eurozone because sooner or later Greece (and in future the other countries who have bankers imposing cuts upon them) will have an election.

    No more austerity and no more shared currency, unless somebody in the centre changes his/her/their mind.

    The scale of the riots already is frightening.

  43. @SocalLiberal

    “Who was Keith Moon again?”

    Keith Moon was a madcap drummer with one of the seminal bands in British Rock history, called “The Who”. Roger Daltry, John Entwistle , Pete Townsend and Keith Moon were in the original line-up and they produced some classic songs in the 60s like “My Generation”, “Substitute”, “Pinball Wizard” etc. The 70s saw them at the peak of their popularity and they became legendary figures on the stadium rock circuit at the time, although Townsend, the main songwriter, was getting pretensions of greater things and got the band involved in some rather overblown rock opera and concept albums like Tommy and Quadraphenia. The punk revolution in the mid to late 70s consigned them, somewhat unfairly in my view, to the rock dinosaur dustbin.

    I saw the band once in my bohemian student days in 1976 at the Vetch Field in Swansea when they headlined a concert that featured Little Feat and the Sensational Alex Harvey Band. To give you an idea of Keith Moon’s character, as the Who took the stage late on that balmy summer’s night in South Wales, he dedicated the first song in the set to some good friends of his who, he said, were residing close by in a famous local institution, visible from the stadium. He was referring to Swansea Prison!

    Good old Keith, who is sadly no longer with us and neither, alas, is that fine bassist John Entwistle. Sorely missed.

  44. @SoCalLiberal @Crossbat11

    Or, more relevantly to the US, “Who are you?” (theme tune for CSI) and “Won’t get fooled again” (CSI: Miami) and “Baba O’Reilly” (CSI: NY).

  45. @TheSheep

    “Won’t get fooled again”

    A classic Who song. Shouldn’t they play that at the next televised Leaders Debate as Nick Clegg takes to the rostrum??

  46. Valerie,ditto regarding those codes.I had to try three times.
    Perhaps in my case time for new glasses ,but the word in
    italics was very difficult to read.

  47. @SoCalLiberal

    “Now missing precincts from three counties wouldn’t neccessarily change the result. But the fact that the state GOP refuses to count the results from those counties makes me think that Paul probably won those outstanding precincts by enough to win statewide.”

    A tad partisan maybe!

    The Maine counties/precincts are fiercely independent and some have strongly resisted completing their straw polls by the set ‘deadline’ date. Some of those missing precincts are still not scheduled to complete their polls until Mar.

    Counties and precincts were warned both beforehand and during the election process, that if their results were not submitted by the set time, then they would NOT be counted. Polling was meant to have been taking place over that whole week before the ‘deadline’ date. It’s not as if there were huge numbers to count,.5524 actual counts votes plus maybe some 900 ‘missing’ is a max of 6450 across 30 counties, that is say some 17 missing precincts, works out at an average of 53 votes per missing precinct or to put another way about 10 votes per missing precinct per day of polling. Wow what a burden to count!

    There was plenty of time to submit results or even to submit partial results, where the threat of bad weather may have denied voters from getting to the polls at the very last minute. It is not as though the bad weather was not forecast and widely known about. The failure to hold some precinct straw polls by the set date was probably deliberate.

    Some precincts had already very publicly announced that they deliberately intended to delay their precinct polling until after the date the ‘final result’ was due to be announced. Then they wanted to be able to vote, presumably to be able to then see whom to chose as the best candidate to achieve an “upset”. They then seemed to want that announced ‘final result’ retrospectively amended to include their deliberately delayed precinct deliberations. Possibly in the hope of basking in the sun of some small ‘glorifed’ publicity.

    Can you imagine the chaos if such a voting latitude was allowed in the Nov polls. Precincts deliberately waiting until the next President had been announced, before then voting themselves, then expecting to have that announced result possibly changed. Chaos!

    So your implication of somehow ‘Paul’ voters were somehow deliberately disenfranchised by the application of a cut off date is probably a tad partisan and perhaps over-played.

    As for your assertion that Paul would have probably won, that to is highly speculative. Romney had a winning vote margin of 194. One of those counties was slightly pro Romney. Washington county voted for McCain in 2008 by a 5-10% margin and seems to be more pro Paul this time. The third county Aroostock is definitely pro Paul by a narrow less than 5% margin in 2008 and also pro Paul this time around. But to overcome that 194 vote deficit, with only 900 possible votes to be counted (and more than four candidates to choose) from would have been a formidable hurdle. It certainly would stretch the term ‘probably’ beyond its uttermost limit. ‘Maybe’, ‘Possibly’ are perhaps closer, but ‘probably’ – I don’t think so.

  48. R HUCKLE,

    “They need to convert back to their own currencies, so they can devalue their currencies and have fiscal control.”

    Fiscal control? Sure they could devalue, but their inflation and interest rates would ballon. A structured default with EU support would see creditors get maybe 50p in the pound.

    Leaving the Euro might see that fall to closer to 10p with the consequences of that being particularly hard felt in London.

    The hard fact is that the Euro imposed on these countries a need for discipline and sound fiscal management that these nations needed. With hindsight there should have been more effort made to make them take the tough action needed before they joined the Euro.

    Equally the EU and ECB should have been tougher on them once they had joined, but that is the dilemma.

    Critics of the Euro blame it for the problems of the struggling members and warn against the EU dictating to member states. They seem to want to blame the centre for the failure of individual members while not wanting to give the centre power to intervene.

    I suspect that in Germany they would prefer individual states to put their house in order rather than binding EU rules, they just don’t think that member state politicians can be trusted to do it.

    I’ve never been a fan of devaluation as an option as it really tends to be running away from the problem rather than facing it. If Greece had faced up to its problems decades back and collected its taxes and reduced its expenditure to fit it wouldn’t be in the mess it is now.

    I can’t see that running away again, all be it a way to deal with the crisis in the short term, is in the long term interests of Greece or any other country.

    Cllr Peter Cairns (SNP)

  49. JimJam

    “Socal – from over here it all looks good for Obama unless a late Electable compromise candidate emerges at or prior to the Republican convention”

    I could not agree more.

    It is interesting to note that faux news (whose presenters and ‘analysts’ have been split down the middle the last months between Mitt-is-only-one-to-beat-Obama camp and we-need-conservative-purity camp) have begun noticeably in just the last week to talk up a brokered convention and Christie/ Rubio floor-nomination etc.

    The way the GOP nomination race has unfurled- combined with recent economic data- makes a GOP win with the current nomination candidates very unlikely IMO.

    But a late entrance from the populist straight shooter Christie or the telegenic and Latin vote securing Rubio could completely turn the race upside down

  50. Rob & Jim Jam,

    I think we (and it would seem Fox News) are falling into the usual trap of overestimating the likelihood of exciting things happening in politics. A brokered convention would be absolutely fantastic to watch…

    …but realistically, they don’t happen anymore :(

    Would be great if I was wrong though

1 2 3 4