A quick update on the latest voting intention polls. This morning’s YouGov poll for the Sun has topline figures of CON 33%, LAB 34%, LDEM 6%, UKIP 14%, GRN 8% (tabs) – more typical figures than YouGov’s Tuesday poll.

There is also a new TNS poll with topline figures of CON 31%(+3), LAB 31%(-4), LDEM 8%(+2), UKIP 16%(-2), GRN 7%(+2) (tabs). Up until now TNS have tended to release their polls incredibly slowly, often a week or more after their fieldwork was finished, rendering them out of date by the time we see them. Fieldwork for this one however finished on Monday so it’s more timely than usual. The last couple of TNS polls showed Labour leads of 7 points, significantly and consistently larger than any other company. Today’s looks more like other polls. (UPDATE – thanks to those in the comments section who missed something obvious in the TNS tabs that I overlooked! TNS have always weighted by 2010 recalled vote, but they are now weighting by 2010 recalled vote AND 2014 European vote – whether or not this is responsible for their figures coming into line with other companies’s is unclear.)

Finally Ipsos MORI put out their latest Scottish poll yesterday. Westminster voting intentions in Scotland were CON 12%(+2), LAB 24%(+1), LDEM 4%(-2), SNP 52%(nc), GRN 4%(-2). We’ve now had three Scottish polls conducted in 2015. The first one from Panelbase showed the SNP dropping four points and raised some speculation about whether their huge post-referendum surge was fading away again. The second from Survation also had the SNP down, but only by two points and this one from MORI has the level of SNP support holding steady. (On top of that, when tables for the Panelbase poll appeared it turned out that the voting intention question wasn’t asked first, it was asked after a question about whether or not falling oil prices damaged the economic case for Scottish independence, so the SNP fall in that first poll may be a question ordering effect rather than a genuine change)

419 Responses to “Latest YouGov, TNS and MORI polling”

1 6 7 8 9
  1. @Unicorn

    “Nothing but good can come of offering a different perspective on these issues. In what ways do you think my approach might be introducing systematic biases? Where do *you* think the SNP curve is heading?”

    My perspective is quite simple. I don’t try to predict where a given VI is going to be in the future. I tend to say “based on present data”. That might be the easy option, but it’s also the sensible one. Swing back may or may not happen. Some believe it’s inevitable. Others believe it to be a myth. As with all aspects of political arithmetic, I’ll believe it when I see (and only for this point in time).

    As for the SNP curve, it confounds all of us. I was amazed at the support Labour had in Scotland in 2010. It’s almost as if Scotland is now playing catch-up to 2010 in the VI sense. Perhaps the ‘Scottish PM’ had more appeal than I ever gave the idea credit. I could understand the people of Kirkcaldy (and perhaps nearby areas) giving Brown their support, but not Scotland as a whole.

    So fast forward five years, and the backlash has finally occurred, and it would seem that some of it is due to Scots not caring for Miliband too much, and some of it is due to Labour being aligned with the Conservatives in the referendum debate (not to mention Labour’s HoC recent voting record with regards the Bedroom Tax, austerity and so on).

    The SNP’s curve? It will depend on the curves of the main Westminster parties. The latter have to be seen to be acting in the interests of Scots (not necessarily specifically the Scots). Smaller parties only benefit when the larger parties are ineffective.

    If we see more and more instances of a gaggle of English politicians round a table talking about the SNP to an English audience, we can all predict how the VI will shift (both North and South of the border). UK politicians ought to discourage the BBC from engaging in such practices. It will all end in tears.

  2. Thomas Robinson


    Whom the gods would destroy, they first make statisticians.

  3. @ Anthony Wells

    You have almost certainly written about this before, but I don’t understand how you decide exactly which polls to include in your Polling Average calculations. I am happy to defer to you superior knowledge in setting the relative weightings, but why don’t you drop old polls in strict order?

    For example in the current list the oldest poll is one dated Jan 2nd from Opinium/observer, but you have dropped several more recent polls including YouGov/Sunday Times of 9th January and several routine YouGov polls (all of which were included in your list last week). Why do some polls cling on to their place in the list whilst others get ‘relegated’?

    I suspect the answer is something to do with not overloading your list with an unbalanced number of polls from particular companies, but if so then surely that could be handled using your weighting adjustment.

    It would be interesting to know because your averaged are widely cited and – unless I am out of date – are even used as raw material for the Fisher model. I am sure that people would like to feel that their compilation is as systematic as possible.

  4. Statement re Leaders debates:

    Broadcasters have published new plans for TV election debates, including leaders of seven UK political parties.

    The BBC and ITV plan to stage debates involving the Conservatives, Labour, the Lib Dems, Green Party, UKIP, the SNP and Plaid Cymru.

    Sky and Channel 4’s plan to host a head-to-head between Mr Cameron and Ed Miliband remains unchanged.

    The broadcasters said the debates would go ahead regardless of whether any party leader refused to take part.

    “The party leaders will be formally invited to take part in these debates. In the event that any of the invited party leaders decline to participate, debates will take place with the party leaders who accept the invitation,” they said in a joint statement. (my bold)


  5. 7 Parties in a debate? That does seem rather a crowd. It’s going to have to be a long broadcast in order to give everyone a say, and that in itself means that perhaps those points covered at the start could be forgotten by the watching public by the end.

    With so many voices, it’s going to be easy to be drowned out or forgotten unless the speaker has a powerful and resonant message.

  6. @Bramley

    Not sure the Sky / Channel 4 debate will be very entertaining should one of the leaders decline to participate.

    Indeed if the debate rumours are true – Lynton Crosby to advise avoiding debates at all costs and Labour to boycott debates hosted by “One Nation Tory” Jeremy Paxman then Sky / Channel 4 could be broadcasting an hour of Jeremy Paxman getting increasingly irate while shouting at two empty chairs…

  7. @ Lurker

    “Therefore, the seat predictive models are moving in favour to Labour.”

    This is only partly true. It is not clear exactly what you mean by ‘seat predictive’. Do you include here *all* models that calculate and post seat numbers, or just the ones that offer predictions for May alone?

    Election-tomorrow models will clearly not move in any systematic direction other than that dictated by the most recent polling data.

    As his past records show, Fisher’s model has been changing steadily in the direction you mention. However that has now all been disrupted by his recent rejig. Electionforecast hasn’t shown much change to date because – as one of the authors posted here the other day – the regression-to-mean adjustment remains at almost its full weighting up to now. So, the projections are still premised on most of the regression taking place before May. He said that the weights are steadily reduced from 50 days out and so (on the premise of unchanged polls) after that you should start to see a sharp swing to Labour (and Ukip, and the Greens and away from the LibDems – currently still billed to rise by about 4-5 points by May).

  8. Mark Hoban, Con MP for Fareham, to stand down. Majority of 17,092 in 2010 so unlikely to change hands.


    Thanks, that is very informative. I meant those projecting forward to May.


    “Indeed if the debate rumours are true – Lynton Crosby to advise avoiding debates at all costs and Labour to boycott debates hosted by “One Nation Tory” Jeremy Paxman then Sky / Channel 4 could be broadcasting an hour of Jeremy Paxman getting increasingly irate while shouting at two empty chairs…”

    I am sure Paxman could start an argument in an empty room.

    I read that the Sky host alongside Paxman would be Kay Burley, Perhaps they should substitute Paxman for Jon Snow for left/right balance. If both leaders skipped the debate, the presenters could have an argument between themselves.

  10. Bill Patrick

    Looking at Question Time, is there some parliamentary convention that no MP is allowed to discuss a health system other than the NHS and the US system?

    Yes. It’s called Ignorance.

  11. CROSSBAT11

    “Is Populus the new TNS/BMRB?”

    Clearly not, the populous would never vote in such an unsatisfactory way!!!!!

  12. I never watch QT – I actually despise the programme – but it seems odd [ though predicable] for there to be complaints from Northern Britain that topics should not be discussed unless there is a participant from that region of the country.

    Does that mean Northern Ireland can’t be discussed if there are no Northern Irish representatives, Cornwall if there is no-one from there – Barnard Castle if I’m not on the proggy?

    I know the Scots are very, very, VERY special but this is taking special pleading to silly limits.

  13. @R&D

    You didn’t watch it so you can’t appreciate but look on YouTube.
    QT : Next Q
    Questioner : Should the SNP vote on English matters

    So the question was NOT about Scotland but specifically on the SNP.

    The result was a UKIP/LibDem/Con rant about Scots.

    Diane Abbott to her credit was the voice of reason

  14. @ Statgeek

    Thanks for your comment. I always find your charts useful.

    “As with all aspects of political arithmetic, I’ll believe it when I see (and only for this point in time).”

    For the most part this is exactly my own stance but I suspect you are not being entirely candid in making your last restriction (‘only for this point in time’).

    Thought experiment: Your graphs show Tory VIs wobbling around 31-32% for a year or more. Some Tory cheerleader comes of UKPR and says: “Leaked figures show tonight’s YouGov poll has Tories on 50%!”

    According to your official agnostic stance your reaction is: ‘Maybe. Maybe not. We’ll just wait and see.’ Right? You have no prior expectations about evidence you don’t yet have?

    Against this, I strongly suspect that if you were a betting person you’d risk a few quid in wagering that the messenger had got things wrong.

    If you accept that there is any truth in this, then all I am trying to do is to go beyond the here and now and convert those evidence-based constraints into predictions about future support levels and outcomes.

  15. @R & D

    Rant about Scots & the SNP

    But if you YouTube it – you will probably find yourself in agreement with Phil Nuttall.

  16. MrNameless

    NI is one where I can see the logic in having a regional debate with SF, DUP, UUP, SDLP and Alliance included.

    As indeed did happen last time. I keep on pointing to the Wiki article in the 2010 debates:


    which includes full details of the ones in Scotland, Wales and Northern Ireland, but it appears no ever follows my links [sulks].

    As to whether they were readily available, I would imagine that in this multi-channel, youtubed age they wouldn’t be difficult to find at the time in 2010. It would be good though for any future agreement to include ways of finding them easily both up to the election and subsequently.

  17. @RosieandDaisie

    I tend to agree with you about BBC’s Question Time, now a dog-eared and tired old format, chaired by another BBC old boy who is, in my view, well past his sell-by-date. I used to enjoy it in the days of Day and Sissons, and Dimbleby was tolerable for a while, but why do television companies never learn about the law of diminishing turns? They are forever flogging old formats and programmes beyond their natural obsolescence, although it could be that my tiring of Question Time is more a result of me now finding so many modern politicians and commentators mind-numbingly dull and shallow.

    I pine for those great Benn v Heseltine, Healey v Patten and Jenkins v Benn face-offs from the 70s and 80s. I need to be careful about nostalgia and false recall but I seem to remember debates and fierce arguments about political ideas in those days, involving genuine political heavyweights who possessed both gravitas and charisma.

    Now it seems to be political light-weights like Abbot and McVey barking unemployment statistics at each other. It’s only saving grace is that it’s marginally better than Neil’s ghastly This Week programme that follows it an hour later.

    Scant consolation though.

  18. Statgeek,

    On 2010, I think it was more that (a) the combination of the Iraq factor fading and tactical voting meant that Labour got support from the Lib Dems, (b) the Scottish Socialist Party had a spectacular implosion, (c) the Greens were hurt by the loss of importance of Iraq as an issue, and (d) the main opponents for Labour in Scotland are the SNP, who had very little attention paid to them in 2010.

  19. @BP
    Yes the NHS question was a bit pathetic. The rather self-satisfied chap in the audience got away without any challenge to his – shall we say debatable – view that continental European healthcare systems are better than the NHS. You would hope Diane Abbott would have a slight grasp of the subject, but none was apparent from her (or any of the others).


    Good post re re Kirkcaldy, Last night’s council by-election was for a ward in that Westminster seat and shows Kircaldy’s love affair may be drawing to a close.

    SNP came a fairly distant 2nd on STV 1st preferences in 2012, which elected 2 Lab & 1 SNP councillor, but he resigned in somewhat dubious circumstances and SNP won the AV by-election comfortably.

    Oddly, though, both the Courier and the BBC report it as if it were a plurality election, both quoting the difference between SNP & Lab 1st preferences as a “majority”.

    That may, of course, be because Fife Council haven’t yet received the full count information, as they only show 1st preference numbers on their website.

    It will be interesting to see where the 2nd and subsequent preferences went. They may even give some guide to how tactical voting may pan out there in May.

  21. Will the Co-Operative Party be added to the debates? They have 31 MPs.

  22. Incidentally there’s a Parliamentary Briefing Paper on the debates which was published last week:


    The idiotic decision to include the SNP and (particularly) PC will almost certainly come back to bite them with the DUP already considering legal action – and if they get in, the other four will have to as well. I don’t think a debate with 12 leaders can be described as anything but a clusterf*ck.

    The real problem is that the broadcasters haven’t laid down any rules for whom they should invite of not – it seems to be based on who shouted loudest. So if it goes to the Courts they will have to infer some and that will almost certainly mean including the NI Parties or excluding SNP and PC (and possibly Greens and Lib Dems). The trouble is that the broadcaster can’t just do what they feel like in an election period, they have all sorts of obligations in their licences.

  23. 12 leaders, for seven of whose parties 85% of the country will not be able to vote.

  24. Sometimes it is possible to feel genuine sympathy for right wing Scots Unionists, as they despair of the lack of interest that “English Tories” have in maintaining the Union.


  25. couper

    “Rant about Scots & the SNP”

    Why such ruse and silly hyperbole?

    I did nothing of the sort. I wrote a reasoned question and do so again: can nothing and nobody be referred to in any question unless someone is there to be offered right of replky?

    The facts are that:
    [1} Such questioning happens all the time.
    [2] I have never known it be a problem until now, when [and because] it involves the SNP.

    It seems particularly daft to affect such concern when the question was largely about England anyway – the clue being in the word “English” in the quote from your post below:

    “Should the SNP vote on English matters?”

    Or don’t you think that English people should have an opinion on the matter?

  26. OLDNAT

    Re the Tory message, thanks for the heads-up.

    Archived for those without freeviews at https://archive.today/2WiZt

  27. Roger Mexico

    Isn’t the key problem that the London bubble media, like the London bubble politicians, haven’t grasped the idea of devolution?

    There is no need for UK wide broadcast debates. If they could grasp that simple reality, then the problems would disappear.

  28. Considering that the proposed debate format is supposed to be a triumph of Cameron’s tactical negotiating nous, why has he not yet agreed to participate?

  29. Since Anthony has now updated the polling average list I can now look at how the different parties are doing compared with 2014 trends.

    The most notable change is that for the first time since early December the LibDems are polling ABOVE their long-standing trend (17 polls above trend level and just 2 below). For the 3rd time out of four batches, Ukip continue to poll BELOW trends (5 above, 14 below).

    As suspected, it now looks as if the Tory rising blip last week was just the result of a rounding error. They are comfortably back on trend now (12 above, 7 below).

    Labour (10 above, 9 below) and the Greens (12 above, 7 below) both remain securely on trend. The latter is perhaps notable given the rather excited response to some of the recent (edge of MoE) polls. As ever it is much safer to look at polls in the round than on an individual basis as they turn up.

    If it stabilises, the LibDem upturn might just herald the start of one form of the much awaited Swingback. More batches needed before we can get a good sense of what this development means.

  30. OldNat (and others)

    Where UKIP’s Scottish vote is coming from is an interesting topic, though one best addressed through the rather limited polling data we have than cultural supposition. I’ve looked at the the recent polling by Panelbase (Westminster UKIP VI sample 55) and Survation (32) as well as YouGov’s most recent from December (approx 27). MORI only found 8 UKIP voters[1] so I’ve ignored it.

    The interesting thing is that there are very few patterns, even where you might randomness throw up some. The WWC, culturally conservative, older male profile we see for UKIP’s core vote in GB in general simply isn’t there. Panelbase found the vote strongest in 18-34 for example.

    This isn’t because such people don’t exist in Scotland (as Rev Stu found to his chagrin recently). It may be because Scottish Labour is still more culturally in tune with such people[2], while the SNP provides a home for those alienated by the Westminster establishment.

    Where they come from is again less defined than you might expect. Panelbase found 45% had voted SNP in 2011, 18% Con, 13% Lab, 9% LD and the percentages from the other polls are similarish. The picture is one that reflects the Holyrood vote as a whole but with Labour losing less and Con and LD a bit more. So they’re not ex-Labour Orangemen, rather rootless non-cosmopolitans.

    But there’s nothing of the situation of one Party’s vote being hit as overwhelming as with the Conservatives in GB generally. Though if you look at 2010 votes the Con and LD bias becomes stronger and SNP weaker, so there may be voters already disenchanted with the Coalition Parties by 2011 who have moved on from the SNP.

    As you might expect they are mainly No voters, especially in the YouGov polls, but Panelbase has 30% Yes and when I looked at 200+ UKIP Euros voters in four Panelbase polls over the summer, it was 38%. So there could be maybe 5% who want a independent Scotland outside the EU[3] and are prepared to vote accordingly. Which ‘independence’ they prioritise in May will be interesting to see.

    I had to go back to the YouGov from the end of October to get cross-breaks by place of birth, but they showed no real difference in UKIP voting: Scots 5%, rUK 7% and ironically born outside UK on 12% (sub-sample only 49). There seems little regional variation either though based on hardly any data.

    In other words Scottish UKIP voters seem to be a fairly random lot, less fanatical Brits than generally discontented. If there is solid demographic for them to exploit they haven’t managed it yet.

    [1] This might be a ‘shy UKIP’ factor or online polls usual bias towards UKIP or both.

    [2] Whether SLab are in tune with anyone else is another question and perhaps a big part of their problems.

    [3] The Wings poll actually found that 2011 SNP voters were the only group of main Party supporters that wanted to leave the EU (by 42% to 38%) and there was no real difference between Yes and No voters on the question (both slightly for staying).

  31. I might have thought about flying my flag at half mast for the deceased Saudi king – but the winter storms beheaded the flag pole.

    That seems much more appropriate.

  32. Often on QT they discuss Ukip policy and Farage/senior member comments, when there is no ukip representation to defend the attacks. How is that fair either?

  33. And the Lib Dems have kicked off their really bad infographics again:


    Note that Con (19%) and UKIP (11%) should be between Lab and Lib. Sigh!

  34. Roger Mexico

    Thanks for the analysis. As with most issues in Scottish polling at the moment is not so much who votes for which party, but where they will vote.

    If Scottish UNS is a useful concept (which I’m dubious about) then we have a reasonable picture of which seats would change hands on current polling.

    If VI for some/all parties is geographically concentrated, then the seat numbers could be very different.

  35. Guymonde,

    One point that no-one made is that we have never had an entirely public system, and that the mix of a public system and private insurance has existed since the earliest days of the NHS. So strictly everyone in that part of the discussion was inferring from false premises.

  36. 7 way debates are a farce, there will be too many voices all talking over one another and they will be a turn off.
    SNP’s and PC’s, main rivals are Labour, and the Greens are left wing and will win most support from Labour. This 7 way debate will not be good for Labour! Even the Tories realise they can’t out UKIP, ukip, so need to attack Labour and the LD’s again will want to win back support lost to Labour.

  37. Another fact that no-one mentioned is that among major Western countries we and the Kiwis are the lowest spenders per head on healthcare. The NHS’s challenges are very much long-term rather than immediate.

  38. @Unicorn

    For your trend line, are you using a form of regression?

    The CUSUM I know and love doesn’t use a regression line, but a simple mean as it’s base.

    It would feel wrong to mix a regression line (with a potential error and inaccuracy), with a CUSUM system (aimed at detect variation below 1.5 SD).

    How is that fair either?

    It’s not. Has anyone written that it is?

  40. @R&D

    They were ranting not you and if you YouTube it you will find it was a RANT.


    I don’t think QT ever have a question specifically about a party without a representative of that party on the panel. I have never known them to.

  41. Statgeek,

    I’m surprised anyone is surprised. What I AM surprised at is how brazen they’re being about targeting Tories and saying Labour’s only just behind – if I was running the Lib Dem campaign I’d want to keep Labour’s challenge as quiet as possible so left-of-centre voters continue to back Clegg trying to keep the Tories out.

  42. @ CMJ

    The approach I am using emerged from exchanges we had some weeks ago (and at the time I acknowledged your contribution to the process).

    I don’t use CUSUM but – if you recall – you persuaded me or its merits. I’d be interested to hear whether you can pick up any LD upturn. (My result is not exactly an upturn, but rather a shift above their small, steady decline.)

    I settled on my rough-and-ready trend-checking method after seeing you post that six (Labour VI) in a row signalled a reliable uptick in support. It seemed that there was an underlying binomial logic to your comments.

    My method is simply to take my Jan – Nov (linear) regression fits for each of the parties and then use the equations to predict where each of the VIs should be in each of Anthony’s successive polling average batches. Then I just count how many of the polls are above or below the projections (and use binomial probabilities to assess reliability). Parametric methods might be bit more precise, but I’m only really trying to provide an early warning system to detect departures from established trends.

    So far, all VIs have behaved surprisingly well – sticking closely to the script, and we are still on target to hit the May VI projections I posted in November.

  43. @ Unicorn

    You keep on making the same methodological error. There is no individual party trend deriving from VI figure simply because it’s a zero sum thing. The only thing you can do is analysing distribution, that is how the adjusted 100% is distributed among the parties. It has nothing to do with what you conclude from the data.

    (You also misunderstand or rather ignore churn. AW summarised it very well how it works. It doesn’t affect ONE party).

  44. @Unicorn

    “According to your official agnostic stance your reaction is: ‘Maybe. Maybe not. We’ll just wait and see.’ Right? You have no prior expectations about evidence you don’t yet have?”

    Pretty much, but my belief levels are generally in tune with norms and outliers, as are most peoples’ belief levels.

    “Against this, I strongly suspect that if you were a betting person you’d risk a few quid in wagering that the messenger had got things wrong.”

    I’m not a betting person (occasional Euromillions if I happen to be in a lottery shop on a Friday). By and large I would privately say “Aye right!” and publicly say “Are you sure? That seems a little unusual.”

    @CB11 (RE: Question Time)

    Completely agree on all points.


    And when someone is present they do not. Quite cowardly, no?


    “and shows Kircaldy’s love affair may be drawing to a close.”

    Not so fast. Let me give you some of the street names in Fife.

    Bevin Place – Rosyth
    Keir Hardie Street – Methil
    Adamson Avenue – Kirkcaldy
    Wilson Avenue – Kirkcaldy
    Lansbury Street – Kirkcaldy
    Henderson Park – Windygates

    Fife will always tend towards Labour as the default option in my opinion. Just as some areas of England always default towards the Conservatives. It’s part of their political roots. Heavy industry until the 1980s, so it will be the 2040s before people start to forget.

  45. “Fife will always tend towards Labour as the default option in my opinion.”

    Meaning the south and west of Fife.

  46. I’ve always thought that the format of QT should be reversed. The panel should ask questions of the audience :)

  47. Actually, now that I think about it, just the south.

  48. Statgeek

    You missed out Gargarin Way in Lumphinnans for the Communists and Jamphlers Road (Champs de Fleurs) in Lochgelly for the Marian legacy. :-)

    The demographics of south Fife, like West Lothian, have been changing in this century as it becomes more of a commuter area for the capital.

1 6 7 8 9