As usual for a Monday we have three GB polls today – Populus, Ashcroft and YouGov. In a election campaign that has so far seen polls that are virtually static these were awaited in the hope they’d shed some light on the impact of the Paxman interviews last week. In the two post-Paxman polls at the weekend YouGov had shown a larger Labour lead than usual, but ComRes had shown a larger Conservative lead than usual. The question was whether today’s polls would shed any light on whether there was any movement, or just normal sampling error.

Populus’s twice-weekly poll has topline figures of CON 34%, LAB 34%, LDEM 8%, UKIP 15%, GRN 4% (tabs). Populus have typically been showing a small Labour lead in their polls over the last few weeks, so this is more Tory than their average poll, but well within the normal margin of error.

Lord Ashcroft’s weekly poll has topline figures of CON 36%, LAB 34%, LDEM 6%, UKIP 10%, GRN 7% (tabs). This is a small shift towards the Conservatives since Ashcroft’s poll last week, but a two point lead is very much in line with the average of his recent polls, so is nothing to suggest any real movement.

YouGov’s daily poll for the Sun has topline figures of CON 35%, LAB 35%, LDEM 8%, UKIP 12%, GRN 5% – back to more typical figures of neck-and-neck.

Looking at the five polls conducted since the Paxman debate, things are starting to look much more like “no change” that a Labour or Conservative boost – there is a bit of movement in either direction, but no clear consistent trend. The seven way debate this week may have more impact, if it’s not just a complete mess.

Note however, that a lack of change in voting intention figures doesn’t necessarily means the interviews last week had no impact at all. YouGov’s weekend poll also saw a significant improvement in Ed Miliband’s ratings and this was echoed in Lord Ashcroft’s poll today. While David Cameron still led on most measures, his lead over Miliband had dropped across the board since Ashcroft last asked in February: Cameron’s lead on representing Britain abroad was down 8 points to 28, on making the right decisions when they are unpopular down 6 points to 23, on having a clear idea of what he wants to acheive down 8 to 19, on leading a team down 6 to 30, on doing the job of Prime Minister down 5 to 26. Miliband’s lead on understanding ordinary people rose 8 points to 12. Of course it would be wrong to necessarily put this down to the interviews, there were signs of improvements in Miliband’s ratings in polls before last week, but it does look as if he’s narrowing Cameron’s advantage.

Meanwhile there were also Wales and London polls out today. The latest Welsh YouGov poll for ITV and Cardiff University has topline figures of CON 25%(nc), LAB 40%(+1), LDEM 5%(nc), Plaid 11%(+1) UKIP 14%(nc), GRN 5%(-1) – Roger Scully’s analysis here. A new ComRes London poll for ITV London has topline figures of CON 32%, LAB 46%, LDEM 8%, UKIP 9%, GRN 4% (tabs).

Note that despite what you may be seeing on Twitter, there is NOT a new ComRes Scottish poll – it’s just people getting excited over a small sub-sample of 70 spitting out the sort of strange and outlandish results that are inevitable with small sub-samples of 70 people.

369 Responses to “Monday polling round-up”

1 6 7 8
  1. Damn HTML confusing arithmetical signs with commands!

    Lab lead in 2010 (smallest to largest)

    Less than 30%
    Ochil and South Perthshire
    Dundee West
    North Ayrshire and Arran
    Aberdeen North
    Edinburgh East
    Linlithgow and East Falkirk
    Aberdeen South
    Kilmarnock and Loudoun
    Edinburgh South
    Edinburgh North and Leith
    East Kilbride, Strathaven and Lesmahagow
    East Lothian
    Central Ayrshire
    Lanark and Hamilton East
    Ayr, Carrick and Cumnock

    30 to 40%
    Edinburgh South West
    Glasgow South
    Glasgow North
    Cumbernauld, Kilsyth and Kirkintilloch East
    Dumfries and Galloway
    Glasgow Central
    Airdrie and Shotts
    Paisley and Renfrewshire North
    Dunfermline and West Fife
    Glasgow East
    Glasgow North West

    Greater than 40%
    West Dunbartonshire
    Paisley and Renfrewshire South
    East Renfrewshire
    Rutherglen and Hamilton West
    Glasgow South West
    Coatbridge, Chryston and Bellshill
    Kirkcaldy and Cowdenbeath
    Glasgow North East

  2. @Robert

    It’s quite obvious to me, judging by the likely seat count, that Labour led Government supported by the SNP is the most probable outcome.

    There is no practical reason against it. It is constitutionally perfectly okay. Before 2010 I’m sure you would have got short shrift proposing a Con LD coalition.

    In the end it’s down to bums on seats in Westminster after the voters have done their duty

  3. One of the good things about this site is that – by and large – people don’t just predict things they hope for. They predict things they think they have arguments for. So the discussion is more thoughtful.

  4. Prof Howard

    The notable difference between the 3 groups of seats is the Green share in the middle group (includes Edinburgh SW).

    The combined share of SNP & SGP (not that they are the same, but it makes the point) are –

    45% : 46% : 46%

    If that Green vote in the middle group is concentrated in a few of those seats, then there wouldn’t be a lot of difference in the other 37?? ex-Lab seats

  5. Average Con VI share in the 3 groups of constituencies (2010 vote in brackets*)

    14% (16%) : 13% (12%) : 10% (10%)

    No sign of Con tactical voting there.

    * Committing the cardinal sin of averaging percentages.

  6. The May 2015 New Statesman site seems to have reversed the poll figures for the two main parties from last weekend’s YouGov Sunday Times Poll. It looks like that is feeding through into the seat predictions as well.

  7. @Raf

    I think Election Forecast UK will turn out to be the most accurate seats predictor. At present they have 285/279 (Con/Lab) on their nowcast and 287/275 on their predictor.

    I agree that EF has the most accurate database. But their regression-to-mean assumptions are open to question and my comparisons with Ashcroft polls has indicated a reliable Labour-favouring bias (at least in constituencies that have not previously been polled). As a result, I think that other models may turn out to be more accurate on the day: perhaps even one of the BES projections .

    More immediately I think you give questionable figures for their Nowcast (a term the stopped using s couple of months ago). The first thing to say that their projections have never been based on a straightforward tally or the number of constituencies in which a particular party is ahead in the seat-by-seat predictions. Rather, the convert the VIs first to probabilities of winning and then they aggregate these probabilities to get the projected seat tallies.

    I have not taken the time to go through this process for the Nowcast,,but if you tot up [1] today’s wins (as a rough approximation) you get 280.5 for Labour and 272 for the Tories – which is much more to Labour’s advantage than the figures you quote above.

    [1] today’s figures show about 14 dead heats for one of the major parties. In these cases I have allocated half a seat to each of the tied parties.

  8. That 37 – 43% Scottish poll makes it look a lot closer than I’d previously thought.

    If these are the seats that are likely to change hands – Labour only need to reduce that 19% swing to 16% and who gets the most seats will be on a knife edge. That looks a lot more do-able – or do people disagree?

    It’s probably unlikely, but given the huge swing involved, that margin does look pretty narrow.

  9. Overall, 78% of the (turnout weighted) sample have given a VI.

    In the 3 groups, the figures are –

    83% : 71% : 77%

  10. @MATT M

    That is a much better result for Labour than they have seen in any recent poll I can think of. An outlier or a narrowing of the polls? We’ll have to see but I suspect it will spur the SNP into a frenzy of door knocking since they can hardly count on the media.

  11. New thread

  12. @Peter Crawford

    Electoral calculus, though not a predictor, is also sensible.

    On what basis do you judge the Electoral Calculus seat counts to be ‘sensible’.

    I would argue that it is best to take their projections with a pinch of salt. My main reason for taking this view is that their projections are so far out of line with the evidence available from Ashcroft polls. When a new batch of polls comes out EC typically shows a pro-Labour [1] and anti-LD [2] and anti-Green [3] bias and even now it is assigning to Labour many of the seats that Ashcroft polling places with other parties.

    Why do you treat as ‘sensible’ a model that gets almost everything wrong?

    [1] For example if the December batch of 13 constituencies, EC overstated Labour,s VI all seats other than Ealing Central & Acton. This is represents a statistically reliable overstatement of their VIs.

    [2] In the same batch EC undestimated the LibDem VI in every single seat

    [3] the Green VIs were also understated in every single seat.

  13. To those proposing that Farron would be the next leader of the LD’s, I recall a discussion a week or so ago, when several were of the opinion that it is far from certain that he will retain his own seat!

    There was a very good article written by Cable, in the Independent yesterday in which he alluded to the fact that a coalition including the LD’s in the next parliament is a distinct possibility. (If the numbers fit) He isn’t pushing for it but neither would he be against it.

    He goes on to say that, ” whatever their differences, the parties (Lib & Con) now have a track record for getting on with it……..the problem with the Labour party has created difficulties in terms of working with them. Quite apart from the economic history, they have a very backward approach to business and their approach to tuition fees (reducing from £9000 to £6000) makes absolutely no sense.”
    He adds ” I am not ruling either out but in both cases (Con & Lab) there are some quite fundamental disagreements. (going forward)”

    So, tough negotiations with both Con & Lab if the numbers fit. Neither is a foregone conclusion & neither is ruled out.

    All that from the LD who was odds on to be the first to quit the coalition within the first year, but who in fact has been Business Secretary for 5 years. The longest since the 1950’s apparently.

  14. ProfHoward

    ITV Scottish poll: has anyone run this swing through the Labour seats to see how many are lost?

    But that’s the interesting thing – the swing isn’t uniform. If ComRes split the 40 seats[1] into three groups depending on how far ahead Labour were in front of the SNP in 2010. If you look at the tables:

    the headline VI results[2] for each set are revealing:

    2010 Lead under 30pts (19 seats):

    SNP 43%

    Lab 35%

    Con 14%

    2010 Lead 30-40pts (11 seats):

    SNP 39%

    Lab 39%

    Con 13%

    2010 Lead over 40pts (10 seats):

    SNP 45%

    Lab 39%

    Con 10%

    Obviously if the swing was uniform the Labour VI would be higher and the SNP lower in each successive group.

    If you applied the swing separately for each group, the result may be that the SNP win everything. Certainly it does hint that Labour may have a better chance of retaining the seats in the middle group. But if you look at these seats:

    Edinburgh South West
    Glasgow South
    Glasgow North
    Cumbernauld, Kilsyth and Kirki
    Dumfries and Galloway
    Glasgow Central
    Airdrie and Shotts
    Paisley and Renfrewshire North
    Dunfermline and Fife West
    Glasgow East
    Glasgow North West

    Ashcroft polled many of them and found healthy SNP leads. So it could be that the fact that SNP and Labour are ‘only’ tied in this group is just sampling.

    [1] With a total sample of a thousand that means that the number polled in any one seat was only about 25. Can we please have no more calls for a breakdown for each constituency.

    [2] Part of Comres’s standard squeeze questioning is to ask what Party people “Generally speaking think of themselves as” and Labour is still ahead in the whole sample here (33%) ahead of SNP (31%), despite the SNP’s big lead on voting. As this answer is used to allocate undecided voters this would help Labour a little. It might also give them some hope for the future, although a lot of people could be answering out of habit and may not return.

    Interestingly only one voter who thought of themselves as Green said they would vote SNP.

  15. Robert N

    DM has the eternal good fortune to have been spared having to prove himself. So he can go down in the Pantheon of Great Leaders Lost.

    Of course, it very much suits the Right to laud the leader that Labour spurned. To ignore the fact that Labour has lost support to its left while DM was the archetypal Blairite. To ignore the fact that the Press ridiculed HIM (not his brother) as Mr Bean when it looked like he was going to challenge Brown in 08/09.

    So, I understand your lauding of DM as the leader who would have swept Labour to victory. But I suspect the reality would have been rather less clear cut than you suggest.

  16. @ Roger Mexico re 5.42 PM post

    Do not forget the ICM/Oakeshott poll from April 2014 that found Cable losing 32% to 34% to Conservative in Twickenham.

    Sutton and Cheam is #9 on the LD marginal list
    Carlshalton and Wallington #29
    Kingston and Surbiton #35

    And according to Lord A all look rock solid LD seats.

    We might be able to recalibrate some of the LD-Con marginals after Lord A’s constituency release tomorrow.

  17. exileinyorks

    “does a poll about the LiS/SNP contest that mentions Lab but not SNP in the recalled 2010 vote seem just a little bit lacking in attention to detail”

    No there may be a good reason for this approach. e.g. It is possibly in recognition that with current SNP supporters there is a lot of false recall. A lot of people who previously did not vote SNP are now claiming they did so. Collecting data for “Other Party” may produce more reliable data and need for less adjustment.

  18. Polls Tonight ?

1 6 7 8