So far we’ve had two new GB polls today, both continuing to show the race pretty much neck-and-neck:

  • Populus’s twice-weekly poll has topline figures of CON 33%, LAB 33%, LDEM 9%, UKIP 15%, GRN 4% (tabs). Note that Populus appear to have made a slight methodology change – their tables include a reallocation of don’t knows, which has the effect of slightly increasing Lib Dem support and decreasing Labour support.
  • Meanwhile a new Survation poll for the Mirror had topline figures of CON 33%, LAB 34%, LDEM 9%, UKIP 16%, GRN 3% (tabs are here)

Elsewhere Lord Ashcroft released his latest, and final, batch of constituency polls – ten constituency polls in a mix of different types of seat. The full details are here:

  • Four of the polls revisited Con-Lab marginals where Lord Ashcroft’s previous polling had found a tight race. In Norwich North (2 point Labour lead), Pudsey (1 point Tory lead) and Wirral West (3 point Labour lead) he found little difference from his previous polls, in Croydon Central he found better news for the Conservatives, with a four point Tory lead compared to a four point Labour lead in March.
  • Another poll revisted the LD-Con marginal of North Cornwall, finding the same two point Liberal Democrat lead as the previous poll in March.
  • Three of the polls were newly surveyed seats – Battersea had been speculated as a seat where the Tories were in trouble, but Ashcroft’s poll found no swing since the election and a twelve point Tory lead. In Stourbridge he found a 4.5% swing from Con-to-Lab, leaving the Conservatives only 2 points ahead. Best of all for Labour was Peterborough, a seat the Conservatives won in 2005, where Ashcroft found a 6 point swing to Labour, putting them 2 points ahead.
  • Finally Ashcroft repolled two Scottish seats. In Dumfriesshire, Clydesdale and Tweeddale (currently the only Conservative seat in Scotland) he found the SNP increasing their lead, up to eleven points from two points last month. In East Renfrewshire, Jim Murphy’s seat, he found the tide moving away from the SNP (possibly due to tactical voting by Tory voters) – in April his poll gave the SNP a nine point lead, the latest poll has them leading by only three points.

Weekly Round up

I’m not going through all this week’s polling like I normally do on Friday’s – frankly there has been too much – but just to complete the record set, the UKPR polling average for the final week of the campaign shows figures of CON 34%(+1), LAB 33%(nc), LDEM 9%(+1), UKIP 14%(nc), GRN 5%(nc), pushing the Tories slightly ahead.

The latest forecasts from Election Forecast, May 2015, Elections Etc, the Guardian and YouGov are below, with the majority of them now suggesting the Conservatives will win slightly more seats than Labour, though only Steve Fisher’s model predicts a Parliament where any sort of Conservative-led government looks feasible.

Elections Etc – Hung Parliament, CON 290(+4), LAB 258(-5), LD 25(-1), SNP 53(+2), UKIP 3(-1)
Election Forecast – Hung Parliament, CON 280(-3), LAB 268(-2), LD 27(+3), SNP 49(+1), UKIP 2(+1)
May 2015 – Hung Parliament, CON 275(+5), LAB 267(-6), LD 27(+1), SNP 56(+1), UKIP 2(-1)
Guardian – Hung Parliament, CON 276(+3), LAB 267(-1), LD 27(-1), SNP 55(nc), UKIP 3(-1)
YouGov Nowcast – Hung Parliament, CON 272(+2), LAB 276(-1), LD 24(-3), SNP 52(+2), UKIP 3(nc)

429 Responses to “Friday polling round up”

1 7 8 9
  1. Guymonde: There has been discussion of May2015 with and without Ashcroft marginals. Just to record that the marginals analysis awards seats away from Lab, but only one of them goes to the Tories: one goes to SNP and 13 to Lib Dem.

    Yes, that’s the net effect.

    May2015 with Ashcroft marginals: Con 275, Lab 267, LD 27

    May2015 without Ashcroft: Con 274, Lab 282, LD 14

    What actually happens is that with Ashcroft, Con gain 9 fewer LD seats, and Lab a couple fewer. But Con gain more of the Con/Lab marginals to compensate for the LD seats not gained.

    So Ashcroft is showing that in some Con/Lab marginals Con are doing better than expected (by national swing). Incumbency or local factors.

  2. Gary O
    Agree about them not using the UKIP for that reason. It would also be dismissed by commentators as no-one realistically expects UKIP to get more than 5 seats at the absolute maximum, whereas SNP will get 10x that

  3. ADAMB and Jim Jam

    It would give a Lab minority a lot more flexibility if Lab + LD seats are higher than Con alone (or Con + UKIP). Far more options for Lab to pass legislation without needing to persuade all the ‘Progressive Bloc’ parties to vote with it.

    Still indicates the same maximum seat difference of around 25 seats.

  4. I know some people dismiss bookies odds however odds on DC remaining as PM are continuing to tumble.

    After across the board shortening of odds post the Thurs debates, there’s more this morning and you now have 10 bookies giving evens against him being PM with 3 even shorter. All still have EM as favourite but there’s now a whisker between them, whereas EM had a reasonable gap a couple weeks ago

  5. Casclc
    I just cant see LD going with Lab, unless Clegg loses his seat. He’s been pretty damning of EM and has said he’d first try to partner with hte largest party, which is pretty much nailed on as the Tories

  6. On of the most laughable trends over the past few days has been the sudden rise of posts by Tory …. supporters quoting “substantial”, “significant”, and “definite” swings in betting odds in favour of the Tories.

    Well, if the polls aren’t doing what you want them to do, and you’re getting desperate to gain momentum, you just might find some solace there.
    But the bookies do not have a good record in getting election results right.

    Post on, brave cyber Cons!

  7. Adamb – I’d ignore the bookies odds – it’s just an indication where the money has gone recently and what’s the value bet. The bookies odds were so far off in 2010, as well as countless other elections, that this market is almost worthless as an indicator

  8. Russell
    I dont place a huge store in bookies, as their odds are driven by two factors:
    – bets already placed
    – what they think will happen

    And its only the second of these factors which is interesting. So you cant read too much into it but to dismiss it as irrelevant is naive.

  9. ADAMB

    The odds on PM have moved since Thursday. However, EM is favourite with all of them as far as I can see.

    Difficult to see polls since Thurs as moving things, perhaps some punters are interpreting EM’s “no deals” statements in a particular way.

    In the longer view, we should remember that it was only on April 20th that Paddy Power first put EM as favourite.

  10. Smithy
    Partly agree, though they wouldnt offer long odds on an outcome they see as likely as they’d potentially lose a packet.

  11. @ADAMB

    But that’s exactly the point. On track record, what the bookies thought would happen in recent elections didn’t happen.

  12. ADAMB: I’m not talking about ‘going with’ Labour, except on a vote by vote basis. I would expect LD to vote with a Lab minority on policies they are agreed on. It could be a very different world to the one we’ve recently been used to.

  13. NumbrCrunchrPolitics tweeted:

    “One bookie taking “large amounts of cash” on CON majority at 11/2, having taken £30k at 7/1. @MattSingh_ has his theory as to why… #GE2015″

    “And no, it has nothing to do with inside information about forthcoming polls…”

  14. SMITHY:

    Exactly. From NumberCruncher yesterday On the betting markets, bookies reported substantial bets going on a Conservative majority, with the odds shortening accordingly.

  15. I understand there is a big story coming out today about Labour , which may affect the polls.

    sorry it was a ( bad ) joke .

  16. Bookies make their money on a spread of bets, the election result is one aspect of this spread which covers a wide range of topics from sport to the sex of a royal baby. Some you win some you lose, but ON BALANCE the bookie makes his money.

  17. JOHNMO

    Thanks for clarifying. I see at least one person on the royal baby thread has entirely missed your joke.

  18. Why don’t we just leave swings in bookies’ odds to betting blogs and consider polls, polling swings and polling methods on a polling blog.

  19. A thought about the idea that there may be a change underway from government by a single (inevitably large) party to government by a block of (potentially small) parties. A large party will likely have its own internal factions, and there may be quite bitter struggles between these, but the public may be mostly unaware of these struggles and largely unable to influence their outcome. If any struggles between different elements of the government are between separate parties, and are more public, then elections provide the public with a chance to pass judgment.

    Perhaps having two large parties competing for an overall majority is not good for democracy, as it reduces the choice for the vast majority of the electorate who are not willing to join a political party, let alone become an activist involved with the internal workings of that party. In any case, I am sure that most people who frequent this site would find a simple two horse race between the main parties much less interesting.

  20. CASCLC,

    I agree as there will be some issue which the SNP would want to abstain or rather be seen to abstain on whilst supporting due to EVEL sensitivities for example on and if LD+Lab was higher than Cons+UKIP it would pass (NI parties?).

    Also the LDs have bigged up ‘national interest, stable Government’ and should Lab lead a Minority Government but the NNP decide to abstain on a confidence motion the LDs may judge that keeping Labour in power is right for the country (and for cynics for them).

  21. On May2015 if you run the poll of polls tool for the last month up to and including yesterday it gives both Lab and Con 33

  22. casclc


    Exactly. From NumberCruncher yesterday On the betting markets, bookies reported substantial bets going on a Conservative majority, with the odds shortening accordingly.

    People make bets when they believe that they know something that you don’t, that does not imply they do.

    Could be an insider, who has spoken to someone with, about private marginal seat polls, that is not public information, and these betters believe it to be true.

    Like I have said it may or may not be. It could be Casino player, liking the odds, should a win occur.

    “It could be even Miliband, making bets on a Conservative win, trying to hedge a life after politics, should he loses. LOL

  23. Looks like ICM have a Sheffield Hallam poll coming up. They’ve just phoned me to ask my VI. I had to tell them we’re 50m on the wrong side of the boundary.

  24. Seats Lib Dems will probably lose

    SNP (9)
    Ross, Skye, Lochbar
    Caithness, Sutherland and Easter Ross
    East Dunbartonshire
    Edinburgh West
    Argyll & Bute
    East Fife
    Aberdeenshire West

    CON (11)
    1 of Berwickshire/St Ives
    St Austell
    Mid Dorset
    Berwick Upon Tweed
    Portsmouth South
    Devon N

    Labour (9)
    1 of Sheffield H/Bermondsey/Birmingham Yardley
    Bradford E
    Manchester Witherington
    Brent C
    Hornsey & Wood Green
    Norwich N
    Cardiff C
    Bristol W

  25. AdamB

    ‘I dont place a huge store in bookies, as their odds are driven by two factors:
    – bets already placed
    – what they think will happen

    And its only the second of these factors which is interesting. So you cant read too much into it but to dismiss it as irrelevant is naive.’

    The second factor is exactly what the bookies do NOT do.

    That would mean they were in effect themselves placing a bet, something they NEVER do.

    Bookmakers do set initial odds on a market on their expectations but thereafter (and in politics case this means over 99% of the time) it is set by following the money.

    It is their living and they always try to set a book that makes them maximum profit in every eventuality, the result is a matter of supreme indifference to them.

    And yet the betting markets are a sort of poll – but a very imperfect and undoubtedly exceedingly biased one, reliant on those that both have the money and the inclination to take the risk of gambling it. A limited section of society exercising their opinions with differing “block” votes according to finance. If there was an actual poll ever done on even partially of this basis you would dismiss it out of hand as unreliable and unacceptable.

  26. On the subject of bookies, after reading many posters’ considered analysis of the polls, I thought 5 to 1 on the LD’s getting 11-20 seats seemed like a good bet, so I took it. Does anyone have an opinion (on the basis of evidence) that I made a wise or poor choice? Thank you.

  27. Jonathan,

    I am a bookmaker.

    We rarely balance our books. We study hard and set odds with the intention of obtaining ‘value’.

    If you think that Man Utd winning or not winning today is of supreme indifference to bookmakers you do not understand our industry.

  28. I think the Lib Dems will lose 10 seats to the SNP (with the 11th reduced from their safest seat in the country to a marginal) 10 seats to Labour, and 12 to 14 seats to the Conservatives, leaving them with 25 to as few as 23 seats. If the students had not had to register individually, they would be losing another 5 seats on top of that.

  29. Sometimes I despair of the BBC interviewers!

    Nick Clegg being interviewed warned about the SNP tail wagging the Labour Dog, if they entered into coalition, and that they would be introducing their own policies, yet they then didn’t ask him about his own time in coalition and whether the Lib Dem tail wasn’t wagging the Tory dog !

    Such an obvious question begging to be asked yet it wasn’t

1 7 8 9