Post-debate polls

We now have all four of the post-debate polls. Asked who performed the best, the verdicts are

YouGov: Sturgeon 28, Farage 20, Cameron 18, Miliband 15, Clegg 10, Bennett 5, Wood 4 (tabs)
ICM: Miliband 25, Cameron 24, Farage 19, Sturgeon 17, Clegg 9, Bennett 3, Wood 2 (tabs)
ComRes: Farage 21, Cameron 21, Miliband 21, Sturgeon 20, Clegg 9, Bennett 5, Wood 2
Survation: Cameron 25, Miliband 25, Farage 24, Sturgeon 15, Clegg 6, Bennett 3, Wood 2 (tabs)

The pollsters certainly don’t show a single clear winner (until we digest the tables we can’t really speculate about what caused the differences), but there’s a consistent theme across those results. Everyone has Natalie Bennett, Leanne Wood and Nick Clegg trailing, everyone has Ed Miliband and David Cameron doing pretty much evenly and pretty close to Farage. The only big difference is Sturgeon, YouGov have her out in front, others have her a bit further back. But given the SNP are only standing in Scotland, how well Nicola Sturgeon did in rest of Britain doesn’t necessarily make a huge difference (though it may make the SNP a bit less scary as a bogeyman on Conservative posters).

As I wrote before the debate, one shouldn’t assume too much what the impact of the debate is on voting intentions until we see some standard voting intention polls of the general public. With the initial polls showing no big runaway winner from the debate I wouldn’t expect a big impact… but time will tell. The party I’d probably watch is UKIP – Farage wasn’t a big winner, but he outperformed his party’s national support in all of the instant verdict polls, and unlike the SNP, UKIP have candidates across GB that people can vote for.

264 Responses to “Post-debate polls”

1 4 5 6
  1. @Rich

    But is the ‘left of centre’ stance the problem for the Tories, vis-a-vis the SNP? I think not. The problem the Tories have is that the growth of the SNP puts the Union at risk, if not immediately then long term. The disintegration of the Union will not harm them in terms of seats (the most they can hope to win in Scotland in this GE is 4), but it will harm their image.

    If the Tories and the LDs really cared about the Union would they not refuse to stand in all Labour held (2010 – 2015) seats so as to ensure the SNP has no chance of winning? Idem mutatis mutandis for LD seats and the one Tory seat?

  2. Rivers10



  3. @DiF – I think what @Rivers10 is saying is that 19/20 Con losses to Lab is nailed on (he/she says) with more seats likely to fall.

    Assuming Cons win 5-10 LD seats, this puts Cons in the low 290’s, with LD ~ 30. Net losses to the coalition parties of 40+ seats means a three party arrangement would be needed to guarantee a stable government, so just 25 Con losses to Lab – highly likely according to the current polls – means a stable Con led government is difficult, even if Lab is back on 265.

    It would then be difficult also for Lab, but because there would be a larger block of SNP votes, a Lab led 3 party group (with LD’s) could be more stable than a Con led 3 party group, even though Cons are 30 seats ahead of Lab.

    Every seat beyond this 25 that moves from Con to Lab has a double impact. 5 more Lab gains (from Con) means that even with the DUP, Cons would struggle for a majority, while the Con/Lab gap reduces to less than ~20 seats.
    So 30 Lab gains from Con effectively stumps Cons from government.

    If Lab can gain 5 more seats off LD than Con (as seems likely at present) and if Lab can hold SNP losses to 20 – 25, then Lab comes very close to largest party, alongside Cons need for a fourth partner for a majority.

    The other political point to consider is that I don’t believe LD’s will necessarily go for the largest party. They will certainly talk to the largest party first, but they will also be looking for ‘strong and stable’ government. Cons + LD would give this, Cons + LD + DUP might give this, but Lab + SNP + LD could also provide this, with a bigger majority likely.

    Cons really need to limit their losses to Lab to 25, maybe 30, to be sure of putting together a new government. Losses of 25 – 40 puts everyone into a dog’s breakfast parliament, but if Lab gain more than 40 from Con, it would look fairly reasonable for a Lab leadership, although this would still be very difficult.

  4. @ Alec

    “They will certainly talk to the largest party first,”

    This is a massive assumption – that anybody would be willing to listen to their voice. Unless there is some big change in the polls (or rather in the actual election results vis-a-vis the polls), they will be pariah.

  5. Its still Lab and some sort of SNP deal favourite, even if Cons get 10-15 more seats. Second favourite is probably Cons staying with Libs. I still don’t see how we will get any sort of strong Govt though. Lab SNP is scary to me as Barnett formula is already unfair, and the SNP won’t be doing a confidence and supply deal for nothing…

  6. @ Rich

    Labour really doesn’t need a deal with SNP – if they can iron out that no Trident and alike. Thus depriving SNP to bring down the Labour government on a moral ground (ie. we would have liked to keep this Labour government in, but they crossed the red line …)

    The most I can foresee another independence referendum and the introduction of some sort of PR (the Labourite who were vehemently against it are disappearing).

  7. @Laszlo – “Unless there is some big change in the polls (or rather in the actual election results vis-a-vis the polls), they will be pariah.”

    Believe me – if the result does end up looking like this, no one – but no one – with a seat in the commons, will be a pariah. They will all be everyone’s best buddy, until the decisions are made.

    Lib Dems are, in my view, likely to be the most wooed block. They may still have a significant number of seats, and unlike UKIP and the SNP they will readily tip either way. The Lib Dems remain as the likely kingmakers, unless Lab get 40 ish gains from Con and make the Lab/SNP block bigger than 326.

  8. @Richard

    If we split the too close to call 50/50, with change from 2010 in brackets
    Conservatives 42.6 (-9)
    Labour 9.1 (+7)
    Lib Dem 3 (-1)
    UKIP 3.1 (+3)

    Interesting analysis. By way of comparison the EF Forecast projections [1] today are:

    Conservatives 47.8
    Labour 6.5
    Lib Dem 3.2
    UKIP 0.5

    This includes shifts resulting from reversion-to-mean, and their Nowcast would be more Labour favouring than your figures above.

    Rather surprisingly, their projection has Carswell’s chance of holding Clacton at only 0.52.

    [1] These figures are obtained by summing over the seats the probability of winning rather than merely counting the number of projected wins.

  9. New post, so I’ll repeat this there. But I wonder if the positive UK-wide reaction to Nicola Sturgeon has a much wider political significance than this election.

    The Lib Dems used to offer a left-of-centre alternative to Labour which had significant appeal because Labour was/is perceived – rightly or wrongly – as coming with too much baggage: being beholden to the unions; wanting to level down rather than level up; too centralist and controlling; guilty of complacent machine politics where locally dominant; class-based. And so on.

    That, I think, is a large part of the SNP’s appeal in Scotland: it offers left of centre policies without any of those negatives. And the abdication of that ground by the Lib Dems leaves a vacuum in England which sooner or later someone is going to fill.

    I just wonder if that ‘someone’ could be an English sister party of the SNP? If the SNP has 40-50 MPs next month, what’s to stop them sponsoring an English National Party (I don’t think the unfortunate right-wing resonance of that name would persist for long, but maybe English Progressive Party would be safer). It could be the SDP all over again, replacing (or absorbing) the LibDems, especially if that party is split between progressives and Orange Bookers.

    As an ex Lib Dem left marooned by lack of a suitable political home, I would welcome that development. And a link with the SNP would work well if the new English party worked towards a federal political settlement, or indeed eventually in an rUK context.

    Clearly there is a widespread dissatisfaction with the current political setup. Could this be the way forward?

  10. @Unicorn

    Thanks for that comparison. One thing that strikes me while going through these Yougov Nowcasts seat by seat is that UKIP’s prospects are a bit brighter than the media narrative that I guess I had bought into. They are running a fairly close second in a number of places.

    That may of course be because I was last looking at the Eastern region where they are stronger. I am currently looking at the North West and they only appear once there.

    I noticed this tweet as well, I wonder if this was also prompted by the release of the Nowcast site 2 days ago?

    SouthHaveringLabour [email protected] · Apr 2
    The threat from UKIP in Dagenham and Rainham is very real. Please contact Jon Cruddas’ office if you can help with campaigning.

  11. @Unicorn

    Note that RAF was also saying that many of Yougov nowcast “the too close to call” seats Lab vs Cons in the London analysis looked like unlikely wins for Labour to him, and it sounds like he knows his stuff, so they may be a pro Lab bias in the Nowcasts. Possibly turnout related – Labour voters being in the demographics less likely to turnout?

    That may also account for some of the large difference between the Yougov Nowcast and Forecast?

  12. @James


    I misreported an earlier regression analysis. The reliable downward slope of 2.4% per month is for the SNP VI alone. (And to answer your later point, this modest drop is still reliable if include the VIs from the four most recent polls.)

    However, to correct what I wrote earlier, the SNP/Labour margin does NOT show a reliable drop since Jan 1. (I mixed up two regression output tables. ). So, while the SNP VI is drifting down slowly and there is a still a tiny numerical drop in the margin, the numbers bounce around so much that the margin drop must be interpreted as being within MoE.

    (It remains the case that the VIs for the sixth fortnight were lower thwn those for the preceding five. But this comparison involved cherry-picking a particular fortnight, and so probably shouldn’t be taken too seriously.)

    Sorry for any confusion cause by my mistake.

  13. Seats like Clacton are so unusual that they have to be treated like outliers (I mean outlier in the true statistical sense). Another such seat is Bradford West. Just chuck any models in the bin for them.


    Thanks for that. I did wonder whether the change was real or could have been due to weighting change(s) as was mooted but you explain it much better.

    In any event, I suspect there will be some more serious Scottish polling soon – perhaps even this week-end prior to the Scottish leader debates, which should provide us with more current information.

    New thread, BTW.

1 4 5 6