Seven Weeks to Go

Here are this week’s polls:

Opinium/Observer (12/3) – CON 33%, LAB 35%, LDEM 7%, UKIP 14%, GRN 7%
YouGov/S Times (13/3) – CON 34%, LAB 34%, LDEM 7%, UKIP 14%, GRN 5%
ComRes/Indy on Sunday (13/3) – CON 33%, LAB 35%, LDEM 7%, UKIP 16%, GRN 4%
ICM/Guardian (15/3) – CON 36%, LAB 35%, LDEM 8%, UKIP 9%, GRN 4%
Populus (15/3) – CON 34%, LAB 34%, LDEM 8%, UKIP 14%, GRN 5%
Ashcroft (15/3) – CON 31%, LAB 29%, LDEM 8%, UKIP 15%, GRN 8%
YouGov/Sun (16/3) – CON 33%, LAB 35%, LDEM 7%, UKIP 13%, GRN 7%
TNS BMRB (16/3) – CON 33%, LAB 32%, LDEM 7%, UKIP 17%, GRN 4%
YouGov/Sun (17/3) – CON 34%, LAB 36%, LDEM 7%, UKIP 12%, GRN 6%
YouGov/Sun (18/3) – CON 33%, LAB 34%, LDEM 8%, UKIP 14%, GRN 6%
YouGov/Sun (19/3) – CON 35%, LAB 33%, LDEM 8%, UKIP 13%, GRN 6%
Populus (19/3) – CON 31%, LAB 34%, LDEM 9%, UKIP 17%, GRN 5%

After a wafer thin Tory lead in the UKPR polling average last week, this week Labour and Conservative are neck-and-neck: CON 33%(nc), LAB 33%(+1), LDEM 8%(+1), UKIP 14%(-1), GRN 6%(nc). The consistent small Labour lead we had for most of the year has vanished, but the Conservatives have not managed to open one of their own.

Scottish and constituency polls

Survation put out a new Scottish poll this week (in fact they put out new – a new one for the Daily Record, and an older one for Unison). Both showed the SNP lead in Scotland holding strong, and no sign of any movement back to Labour. There was also a new batch of Ashcroft polling which I discussed here

Budget Week

This week was the most significant fixed event of the year before the general election. We still have the parties’ manifesto launches and (perhaps) a debate or debates of some sort, but the budget was indisputably one of the most important known unknowns that could potentially change public opinion before the election. The commentariat invariably spend the run up to the budget speculating about what rabbits the Chancellor will pull out of the hat to make a major impact on the election – history suggests it very rarely happens. There are examples of government support slumping after bad budgets – one of the biggest shifts in public opinion this Parliament came after the bodged 2012 “omnishambles” budget – but precious few of them of them giving a government a substantial boost.

We are not yet in a position to judge whether or not the 2015 budget had any effect. The initial post-budget polling from YouGov gave it a thumbs up – more people thought it was fair than unfair, more people thought it would make them better off than worse off. Most however thought it made no difference, and even a positive reaction does not automatically lead to any impact on the polls. The two voting intention polls we’ve seen since the budget do not point to any big shift – YouGov showed the Conservatives up, but Populus showed them down.

Beyond the budget there was another tiptoe forward in the debate debate. The broadcasters are reportedly now offering one seven way debate on ITV, a Paxman grilling of Miliband and Cameron on Channel4/Sky and two Question Time style events on BBC – one for UKIP, the SNP, Plaid and the Greens, one with Miliband, Clegg and Cameron, but separately. Cameron has agreed to the seven way debate, but all else seems unconfirmed. The first event – the Paxman grilling – is pencilled in to take place next Thursday though, so one way or another, decisions will be made next week.

Projections

The latest forecasts from Election Forecast, May 2015, Elections Etc and the Guardian are below, as well as the new model from the Polling Observatory team. As usual all the models predict a hung Parliament – most have the Tories with marginally more seats than Labour, Polling Observatory have Labour with a 20 seat lead.

Elections Etc – Hung Parliament, CON 284(-1), LAB 278(-1), LD 21(-1), SNP 41(+1), UKIP 3(nc)
Election Forecast – Hung Parliament, CON 284(-4), LAB 278(+7), LD 25(-1), SNP 40(-2), UKIP 1(nc)
May 2015 – Hung Parliament, CON 277(-4), LAB 268(+5), LD 24(nc), SNP 55(nc), UKIP 3(-1)
Guardian – Hung Parliament, CON 277(-2), LAB 268(+3), LD 25(-2), SNP 54(+1), UKIP 4(nc)
Polling Observatory – Hung Parliament, CON 265, LAB 285, LDEM 24, SNP 49, UKIP 3


531 Responses to “Seven Weeks to Go”

1 9 10 11
  1. I think Cameron blocking Nick Clegg from one debate is a mistake. It runs the risk of looking unfair and bullying. Could even benefit the LD’s.

  2. I think it is safe to say the outcome of the budget is “no change” in the current polling levels. The Observer poll has the Tories ahead by a few; YouGov has Labour ahead. All within the MoE. It’s probably why the Sunday Times and YouGov aren’t publishing their results with any urgency.

  3. @CATOSWYN

    As I mentioned, I can’t see how Cameron could have ‘blocked’ Clegg’s entry. I think someone is being a bit disingenuous here.

    My feeling is that there has been a bit of horse trading going on – with the LibDems getting representation in the final event in exchange for removing themselves from the 7 way debate (at Cameron’s behest).

    This suits Cameron who wishes to present the opposition as a shambolic farce of bickerers (compared to his aloof regency), but it also suits the LibDems too as being seen as a party of Government.

    I think the LibDems are having us on – they were willing participants in dropping out of the ‘oppositions’ debate, in order to take place in the interviewing programme. To say they were ‘blocked’ is spin on their behalf.

  4. Good Morning as the sun rises here in Bournemouth.
    RORY HUGHES.
    I find it hard to believe that the LD’s would try ‘having us on’.
    Ed M will have to work hard, I think, not to be ‘sucked in’ by the Opposition dog fight which the six way debate session could turn out to be.

  5. Why do you think the LibDems would rise above such a thing, and how do you think that Cameron could effectively block Nick Clegg from appearing in any debate, without his own collusion?

    Cameron has not only managed to squirm out of the debates he didn’t want, he’s essentially disparaged the other parties downgrading an ‘opposition’ debate, sneaked a way of reducing the claim that only he is avoiding debates by taking the LibDems out too, and has sneaked in a three way debate which is essentially a 2/3 government vs 1/3 opposition Q&A set of shows – which are dependent on the quality of the interviewer and his her bias (Kay Burley, anyone?).

    The LibDems would have had to go along with it by allowing themselves to be shifted out of the multi-party debate and into the proposed Con/Lab debate. I don’t believe they are at all displeased with this result at all. They’ve actually got what they want out of it – and actions speak a lot more than words.

    Flatly, I think the proposed outcome is a disgrace to the BBC, the Conservatives and the LibDems.

  6. I’ve just punched the data from last night’s You Gov into my spreadsheets.

    Con

    Another decent poll. That makes it twenty data points out of last twenty one above their 2015 mean since the 20th Feb. Five poll rolling average 33.6, second highest of the year.

    Lab

    An extension of a decent run, with four out of that last five polls above their 2015 mean. Since the 29th Jan Five poll rolling average of 34.6, the second highest of the year.

    LD

    A steady rise since the 18th Feb, now leveling off. A five poll rolling average of 7.6, second highest of the year.

    UKIP

    Down, down, down.

    This path has really headed south broadly since the 19th Feb.

    The last six polls have been below their 2015 mean.Their five poll rolling average is now 13.2, the lowest of the year and the same as yesterday.

    By way of comparison their rolling average peaked at 15.2, in mid Feb.

    Green

    The long run of poor polls continues.

    The 18th Feb was their peak. All data points, except for three, have been below their 2015 mean during this period.

    Current five poll rolling average 6.0, third lowest of the year. On this measure they peaked at 8.0 at the end of Jan.

  7. Correction

    I’ve just punched the data from last night’s You Gov into my spreadsheets.

    Con

    Another decent poll. That makes it twenty data points out of last twenty one above their 2015 mean since the 20th Feb. Five poll rolling average 33.6, second highest of the year.

    Lab

    An extension of a decent run, with four out of that last five polls above their 2015 mean. Five poll rolling average of 34.6, the second highest of the year.

    LD

    A steady rise since the 18th Feb, now leveling off. A five poll rolling average of 7.6, second highest of the year.

    UKIP

    Down, down, down.

    This path has really headed south broadly since the 19th Feb.

    The last six polls have been below their 2015 mean.Their five poll rolling average is now 13.2, the lowest of the year and the same as yesterday.

    By way of comparison their rolling average peaked at 15.2, in mid Feb.

    Green

    The long run of poor polls continues.

    The 18th Feb was their peak. All data points, except for three, have been below their 2015 mean during this period.

    Current five poll rolling average 6.0, third lowest of the year. On this measure they peaked at 8.0 at the end of Jan.

    ——————–

    I thought the budget was one of the set pieces that could really have boosted the Conservatives. Clearly this has not happened, unless it’s a real slow burner.

  8. Good night from here on the west side of Canada.

    Reflecting on the polls over the last week either Lord A’s constituency polling is an accurate finding of what is happening in Scotland, England and Wales or he has seriously mis-read the mood of voters in the constituencies he has polled.

    To date Lord A has found no classical swingback for either Conservatives or their coalition partners the LDs, and in fact a re-visit to 8 Lab-Con marginal constituencies found 6 of 8 gained by Labour, 1 tied and only 1 held by Conservative.

    And yet some pollsters show the Conservatives slowly returning to their 2010 election result. Further Lord A clearly showed that voting intention in these eight seats for LD was 4%, with constituency intention increasing to 5%.

    And yet again this last week not one pollster had LD UK wide below 7% and one had them as high as 10%. There is, I believe, a general consensus that Lord A’s constituency polling in Scotland is basically accurate, even if one does not share the degree to which SNP will sweep Scotland.

    So what happens to Lord A’s polling when it crosses the border to England, does it suddenly become unbelievable for some strange reason. That is what every single UK pollster would have you believe except Lord A, that LD is on a slow rise, a recovery towards the May 7 election.

    And as for UKIP despite repeated constituency polling by Lord A and others, a number of pollsters would have you believe that UKIP is on downward trend and will take few if any seats on May 7th.

    Both points of view cannot be correct simultaneously, and so some pollsters have some serious re-thinking to do in terms methods, modelling and weighting.

    In Scotland the LDs are trying to put a brave face on things by saying they have a fighting chance at six seats and that UK wide they think they can save half their seats. Political parties always put a spin on bad news otherwise how do you motivate the faithful, but pollsters should not engage in wishful thinking and inaccurate reading of the public mood before an election.

  9. Am I correct in thinking that the opposition normally gains support in the last month?

  10. Pete,

    No, the swing back is normally to the Government party.

    David

  11. Interesting article in Scotland on Sunday about a new LiS line launched by Ed in Scotland tomorrow.

    http://www.scotsman.com/news/politics/top-stories/ed-miliband-unveils-offensive-linking-snp-to-tories-1-3725108

    Hard to know if it will work or not in changing oppinion and voter behaviour. I suspect it will only harden oppinion on both sides when a cors vote strategy is not what Labour needs.

    The big issue will be if the SNP claim can be proved, that Labour vote with the Tories more, in which case the whole thing could unravel.

    Given that you are voting for or against the Proposal put forward for your own reasons as opposed to being an alliance of Parties I am not sure it will work, particularly as this is how C&S would work.

    Peter.

  12. Rory,

    You may be right about the ST being in no hurry to release but can’t charge YG with the same.

    YG never post results until the following morning as the client wants to be able to decide when to release and so dictates I think a 6.00 am time.
    Typically weekday ones for these Sun are Tweeted (often Tom Newton-Dunn) especially is it looks good for Cons or bad for Labour or to be fair if it just looks interesting but the ST ones are often not tweeted or put in the public domain early.
    Of course, in London the ST is available around midnight to buy.

  13. For the record, the full YG poll is out and it suggests that the post budget bounce was short lived:

    CON 33 (-2)
    LAB 35 (+2)
    LIB 8 (=)
    UKIP 14 (+1)
    GRN 5 (-1)

    Fldwk 19th-20th

  14. David

    “No, the swing back is normally to the Government party.”

    No, that is not the case. The history of final month swingback is inconsistent, but generally favours the opposition as Pete says.

  15. DAVID and PETER.
    Graham often points out to us that there is normally an improvement in the position of the Opposition during the campaign; possibly this is due to the extra coverage. Lib Dems in particular seem to improve.

  16. Here is this week’s YG data by 2010 voter ID.

    https://drive.google.com/file/d/0BzTTW1ecy-NDcVVsNGgzS1pEUmc/view?usp=sharing

    I have made small methodological change. I used post the changes as is. I then wanted to omit the noise, so changes below 1.96 * SE were omitted.

    More recent analysis of the data by other methods suggested that patterns that should have been visible in the change graph were being omitted as noise.

    Therefore, from today I reduced the filter from 1.96 * SE to just SE. Given the fact that these are very particular datasets, I don’t think I will find an verifiable statistics guide to help out. Accordingly, you will all have to put up with my executive decision ;-)

    Week on Week

    So to the data. Week on week, given the budget was a Conservative set piece, the Conservatives must be disapppointed. As I mentioned before, maybe it’s a slow burner. Only time will tell.

    Labour seem have gained a bit all round, and UKIP have been the losers.

    Week vs Start of Year

    2010 Conservatives

    The shift has been from UKIP back home. The evidence is clear that UKIP have been slipping, and the biggest part is their 2010 Con supporters going home. Maybe the ‘vote UKIP get Labour’ message is working?

    2010 Labour

    This group have drifted back from UKIP and the Greens to the Conservatives and Labour.

    2010 Lib Dems

    Some of the 2010 LDs who stomped off into the arms of Natalie have moved away, and some have left Labour too. Are they switching tactically to give the Lib Dems more options in a hung parliament? This is a bit tricky to analyse.

    2010 Others

    This band is made up of new voters, voters of minor parties and previous non-voters. Labour have been by far benefited the most. It looks like Labour attempts to shore up the anti-Governemnt vote are working to a degree.

  17. Latest yougov

    England 35-35
    Scotland 43-28

    Geddit ?

    Otherwise everyone Is more popular ,economic ratings all going positive,budget approved,noone likes tony blair,snp likely to support labour.

  18. @Andy S

    I share your doubts about current seat projections for the LibDems. As I pointed out to someone yesterday, Electionforecast is still projecting a LD vote share of 13.6%. This is because the reversion-to-mean component of the model is still having quite a strong say. But it beggars belief that over a mere 6.5 weeks a party can raise its VI by something like 60%.

    In the case of EF the weight of the reversion-to-mean will be lowered steadily from now on, and as this happens the number of projected LD seats should drop to the high teens. However (apart from Electionsetc) the other models don’t have access to an automatic correction of this kind, and as far as I can see their LD seat projections will probably remain at the present levels.

    One way for triangulating would be to rank their seats in terms of margin and work through their constituencies ranked (say) from 10th safest to 30th safest and see what mood their is in the comments to be seen in the UKPR constituency section. (Obviously reports from canvassers would help but you are unlikely to have access to such information from your base in Canada.)

    Turning to Ukip though..

    ..a number of pollsters would have you believe that UKIP is on downward trend

    Personally, I don’t question the evidence for a downward drift. It has now been possible to pick this up in a number of different kinds of analysis.

    You have frequently stated that you don’t trust the Ukip VI figures, largely because of the doubts you have about the way some of the pollsters ‘fill in’ the responses of individuals who say they don’t know or won’t tell which way they intend to vote. Well, I think that everyone would have to acknowledge that this could well affect the absolute level of the reported Ukip VI percentage. But – to return to your statement above – it seems highly unlikely that a systematic error of this kind could produce an apparent ‘downward shift’ if support were not genuinely fading. For this to happen don’t know percentages would have to be rising progressively over the period in order for an increasing number of shy Ukip voters to be handed over to another party as part of the adjustment.

    It is because I don’t think this Ukip decline is an artefact that I think that the models are right in projecting that the Ukip prospects are receding and that in the 2015 GE their impact will not be all that great.

  19. EF updated this morning. Changes from Friday update in brackets

    Con 286 (+1)
    Lab 276
    SNP 39 (-1)
    Lib 26
    DUP 8
    SDLP 3
    PC 2
    UKIP 1
    Grn 1
    Oth 8

    Swing from SNP to Con in Scotland! :-)

  20. Suppose one question which remains is what will happen to the Greens now? They’ve been drifting backwards for a month or so, possibly not helped by Bennett’s media appearances, but since then they’ve had no coverage at all.

    I’m unsure if their debate appearance will help them or hinder them. My view is that after Clegg, a lot of people will be very cynically watching the debates and not receptive to the “new broom” challengers, with the possible exception of Sturgeon.

  21. @Chrislane

    Good Morning to you, sir.

    I note your remark above that ” Lib Dems in particular seem to improve” towards the end of the election campaign.

    May I enquire as to whether you are feeling entirely yourself this morning?

  22. Postageincluded

    The Lib Dems are often said to improve in election campaigns because thats the only time they get something close to equal coverage.

  23. OLDNAT

    We don’t even hate the stupendously stupid English folk – but then most English folk pity them as well.

    Do you think that’s funny?

    Try re-reading it and replacing the word ‘English’ with ‘Jews’ or ‘Gays’ or ‘Scottish’ and see if you still thinks it’s funny.

  24. @CMJ

    Your Voter ID Analysis is as fascinating as it always is.

    On the LibDems you write:

    Are they switching tactically to give the Lib Dems more options in a hung parliament?

    Your question seems to presuppose that these voters are LDs in perpetuity just because they voted for the party in 2010. Only on this basis, could you ask whether they are still working out how best to help the LibDems.

    More plausible in my view is that a high proportion of 2010 LD voters are now totally alienated from the party (fees, getting into bed with the Tories etc.) and they are now trying to find a party that represents the principles they thought they were voting for in 2010. Perhaps some of the Green policies don’t quite chime. They may well end up with Labour after a lot of soul searching.

    @Andy S

    To buttress my previous point about the reality of the Ukip decline, it is worth noting that @CMJ’s Voter ID plots are not subject to anti-Ukip biases in don’t know corrections. These plots have consistently shown that individuals who identified themselves earlier in the year as Ukip supporters have now shifted their allegiance to other parties. I just don’t think it is plausible to argue that there has been no recent downward trend in Ukip support.

  25. Howard – will this still apply in 2015 as they have been getting plenty of coverage as part of the Government.

    I post as one who believes they will recover somewhat but due to ABT and ABLabs voting tactically in the end and 2010 LDs saying DK breaking in the LDs favour.

  26. @Unicorn

    You are probably right about the 2010 Lib Dems.

    They are the hardest group to work out, with a mix of motives and complexities.

  27. POSTAGE INCLUDED.

    I could not possibly comment on your question, but I feel well thanks, as the summer approaches, and another school exam season has been survived. The GE campaign approaches, which I enjoy. I also have picked up from twitter two articles, one my the ‘good right’ website and when from Maurice Glasman; about politics, which I can give to my sixth form students.

  28. New thread.

  29. Bluebob

    I think the answer to the question re why some people might think the Tories are competent economic managers but still not vote for them comes down to a couple of things.

    There are a lot of people fearful of more cuts or ‘top down’ reforms to public services. The ‘fairness’ factor is something Labour have tapped into well and the Tories are still not trusted with public services. The Tories had the chance to scotch this in governemnt by investing in the NHS and changing perceptions but they did the reorganisation (rightly or wrongly) and there are also the issues around A&E / Cancer waiting targets / Mental Health / Social Care that dent their credibility on the issue.

    Also, if you and feel that the Lib Dems have been a brake on the most extreme aspects of Tory policy, as some in the centre might, then the thought of a Tory majority without that brake might not be particularly thrilling.

    If the BBC’s piece the other day about the country moving back towards the left whilst the coalition hs been in power is true, then the Tories trying to get the voters to support something more right-wing than they have seen in the last 5 years is always going to be a tough sell.

  30. The EDL allegations are pretty toxic for the Tories. They need to get on top of that story quick.

    They already struggled enough at the last election re votes from ethnic minorities but events like these will nail the Midlands and London swing seats on for Labour.

  31. @pete b
    am i alone in thinking if these predictions are correct or a slight variation,we are as a country in an almighty mess.
    Surely there is going to some form of paralysis in government.Every bill could be stopped or delayed whilst another party tries to squeeze the last concession out of any minority coaltion.in this case surely there will be big trouble for the country and the only likely outcome is another GE.
    We truly seem to be in unchartered territory

1 9 10 11