A quick update on two new voting intention polls yesterday.

Opinium in the Observer has topline figures of CON 39%(-5), LAB 45%(+4), LDEM 5%(-3), UKIP 5%(+3). Changes are since the general election. Jeremy Corbyn’s net approval ratings are now substantially better than Theresa May’s – 31% approve of how May is doing her job, 51% disapprove; 42% approve of how Corbyn is doing his job, 38% disapprove. Full tabs are here.

Survation meanwhile has topline figures of CON 41%(nc), LAB 40%(-4), LDEM 7%(+1), UKIP 2%(nc), the first poll since the general election to show the Conservatives ahead (if the changes since the last poll look odd, it’s because the Tory share is actually up by almost a point, but it’s lost in the rounding, and the share for “other” parties is up three points). Full tabs for that are here.

252 Responses to “Latest Opinium and Survation polls”

1 2 3 6
  1. Yay! The first

    Have Opinium changed their methodology since the election?

  2. I mentioned on the previous thread that, if the Conservative Party are around 40% when they are have as many problems as they do right now, then are are far from doomed.

    A Labour government is far from inevitable.

  3. Are polls other than survation changing methodology?

  4. Sorry tonybtg for posting roughly the same question

  5. Still hung parliament territory.
    The public really doesn’t rust either of them to run the country.
    Also think the Grenfel flat fire might have hurt Labour a small amount…or there poor handling of it after the first week. It needed calm heads not idiots like Mcdonnell spouting off.

  6. Dan

    No worries. It’s just that if opinium have changed their methodology you would expect the two polls to be broadly in line. But it seems we’re still following the pre-election pattern of differing polls.

    If Opinium have a 6% Lab lead on their old methodology, that was showing moderate Con Lead, and were sticking with it. This poll shows a remarkable swing to Lab. But if the methodology has been tweaked then the swing is less.

    If that makes sense :-)

  7. PAUL CROFT, trouble is Labour have similar problems as the Tories. Neither party seemingly knows what to do about anything without having an all out war in its ranks.

  8. The survation poll also found that most people would now vote to remain in the EU.
    Some of the results look odd to me. There’s suddenly a huge gender difference when it comes to remain/leave (page 9), which, as far as I know, has never existed in the past.
    The South of England (minus London) is now much more keen to stay in the EU than Wales (page 18), while London is now way more likely to trust Theresa May than the North does (49% versus 39%).


    Agreed although I will say this. If the Tories remain in power until 2022 I do think the country will have had their fill of Johnson, Gove,May, Duncan-Smith etc and want change.

  10. As has already been remarked, the Survation poll is a bit weird. Every other post-election poll we’ve seen (including from Survation) has shown Labour benefiting from the ‘winner’s bonus’, even though they technically didn’t win. And post-election events seem generally to have amplified that movement. This reverts to the election position:


    CON 41% (+1)
    LAB 40% (-4)
    LD 7% (+1)
    UKIP 2% (NC)
    Others 10% (+2)

    (Comparison with 19/6 Survation. SNP 2%, Green 2%, NI Parties 3% Other 3%).

    Fieldwork is 28-30 June, so Labour’s ‘split’ on the evening of 29th shouldn’t have had much effect. So if the polls are moving against Labour you would expect a similar shift towards Leave and a harder Brexit.

    But on the contrary EU Rerun Ref VI shows one of the largest Remain maybe since the Referendum[1]:
    Leave 46% (-3) Remain 54% (+3)

    and support for “Leave the customs union” drops from 27% to 24%, while “Stop Brexit Negotiations altogether and work to remain in the EU” jumps from 25% to 36%, now the most popular option. Hard Brexit (“Leaving the EU without a mutually agreed deal on the terms of exit would be good for Britain”[2]) falls from 31% to 26% and its converse rises from 58% to 66%.

    So it’s all very odd. One notable thing from the tables is that the raw sample was very Tory with 380 who voted Con last month against only 292 Lab. Obviously this is re-weighted and it doesn’t seem to have been a problem in previous Survations, so maybe they were unlucky. But the anti-Brexit movement needs watching[3] to see is something is really happening as people actually begin to think about the realities of negotiation. Though one odd-looking poll is not a lot to go on.

    [1] It’s difficult to be sure exactly because most such polls have been online rather than phone, and the latter have lower DK responses.

    [2] I know we talk about Brexit rather than UKit, but this seems a little insensitive in a poll that also has NI respondents.

    [3] One additional mystery about this poll is who it was for. Pollsters don’t normally poll for free, though it’s possible they piggybacked on a commercial poll. It refers to “Survation’s EU Trackers” and the previous such poll was for the Mail on Sunday – did they not like the results? Normally the MoS is pro-EU (part of the titanic Dacre v Greig battle) so you wouldn’t expect so. Is it post-election belt-tightening?

  11. Fantasy thinking I know, but if both parties were to split and the two moderate wings formed a new party, one could see the vote share becoming 20/40/20 in favour of the new party. I wonder how that would look in terms of seat share?

  12. Has there been any methodology changes by any pollsters, AW?

  13. Paul croft

    I don’t think anyone is suggesting the Tories are doomed. I don’t think that would be healthy for our democracy anyway.

    The Tories are a party that continually re-invents itself. We’re seeing that process already, with rumours of a relaxation of the pay cap, and calls for the revision of policy towards tuition fees.

  14. ANDREW M

    I think you are speculating on the emergence of Monty Python’s Sensible Party.

    As Buddy Holly sang – “That’ll Be The Day.”

  15. Andrew Myers- I think a new party made up of chukka Heidi Allen etc would sink without trace which is why neither will do it

  16. LAB should be worried.

    Despite the chaos of Brexit and the snap GE – which, to be blunt, were both brought about by CON and to resolve issues within their own party – and despite the DUP deal &c….. LAB are still not 7 points in front of CON. And, in some polls, are still behind.

    Without that lead, they cannot form a majority government.

    To my mind, despite the relative success of the GE for Corbyn and LAB, they need to move sharply towards the centre if they want to capitalise on the current situation.

    Failure to do so will see LAB & Corbyn lose the next election and CON win – with a majority!

  17. Roger Mexico

    Even accurate pollsters will get the odd outlier poll. Wait until the next one. I don’t think there is any need to get too hung up on an individual polling data point.

  18. Two points about polls in the “new world order”:

    1 We haven’t got a clue how accurate they are – if they stick to one methodology they are good to tell us direction of travel (probably!)

    2 We now know that what polls say at the start of a GE campaign can in fact be very different to what they say on polling day

  19. David in France

    “they need to move sharply towards the centre”

    One of their troubles is that the centre is at different places for different policies.

    On tax and spend (on social services) they are probably at the centre. However, on benefit caps they aren’t (I think).
    On nationalisation of railways they are probably in the centre (in London …). On nationalisation of other utilities – the public doesn’t seem to care.
    On tuition fees (as long as financing doesn’t come up) they are in the centre.
    On investment – I don’t think there is much public interest (without much concrete elements, it is not too surprising).
    The only thing where they are not in the centre is Brexit. I’m quite sure that it’s because there is no centre on this issue. Various issues of the negotiations (?) create such trade offs that can easily split both parties and public opinion.

    As to the street theatre, it’s obviously not centre, but unless something goes wrong, it won’t matter.

    So, imo, there is not much opportunity to “move to the centre”.

  20. Regarding the methology of the polls that understated Labour due to youth turnout.

    Is it likely that they have reset the methodology automatically. Because they are now basing it on the actuals who voted in 2017 and not 2015

  21. Paul Croft

    I agree with your first post on this thread and i also think that the latest Survation is much more likely than the Opinium poll which puts Labour much to high, IMO of course.

  22. I’ll personally be letting the polls roll by for a few months before paying attention. I wonder when You Gov will start again?

    More broadly, it seems we in a period where the old economic model that fell apart in the financial crisis is dying, yet the public haven’t got fully behind one alternative convincingly over another offering (thus the parliamentary stalemate).

    There is an archive on four podcast that is very good on this:


    It argues that a radical change needs the message and the messenger that embodies it. Perhaps people feel both haven’t aligned at the moment with anyone or any party.

    Uncertain, but exciting times.

  23. Exactly Matt.

    Opinium, ICM and Coms Res weighted by demographic profile to the actual 2015 turnout while Survation accepted the declared LTV which over stated proportionately hence accurate in vote shares for the big 2 (E&W).

    Accordingly reweighting to the actual 2017 turnout by demographs automatically changes the adjustment to the raw data.

    Most likely higher turn out amongst younger voters and 20-15 DNVs will occur at the next GE as well so the 2017 weightings may be correct enough for the while at least.

    All we can say with any confidence is that a new GE in the next 18-24 months would be close.

  24. David in France
    ‘ LAB are still not 7 points in front of CON. And, in some polls, are still behind.

    Without that lead, they cannot form a majority government.’

    That is very unlikely to be true any longer. From looking at the results in Scotland it would seem that Labour is well placed to pick up a further 25 seats there alone next time . That would take them to over 30 seats – mainly though not exclusively at SNP expense – and Labour would once again be comfortably the largest party in Scotland. The SNP could very easily struggle to win 15 seats next time.

  25. Graham

    Taking seats off the SNP is neither here nor there, all that counts is taking seats from the conservatives. It doesn’t really matter hugely who is taking seats from the conservatives, obviously I rather it was lab but I see there are a few seats where the dems are close and a SNP revival which wipes out the Tories in scotland is very much in our interests.

  26. “TONYBTG @Paul croft
    I don’t think anyone is suggesting the Tories are doomed. I don’t think that would be healthy for our democracy anyway.”

    Well I was using “doomed” in the narrow context of the next election, whenever that may be.

    The “one more push” narrative will only stand any chance at all if Labour does not assume it has won all arguments, internal and external [I think their leadership will assume precisely that] and the Tories don’t get their act together in some way.

    Since they are now more popular than their current leader, instead of the reverse, I find that a very unlikely scenario.

    Mind you, throw in a declining SNP and – possibly – a developing Lib Dem party under Cable, plus the chaos of brexit and quien sabe? [as we say in Europe].

    All in my ever so ‘umble opinion of course.

  27. Graham/CR,

    I think you are both right in that taking seats of the SNP matters little is stopping a Con led Government but it does help get to an OM with a lead less than 7% and most seats with a lower lead than otherwise (have seen different numbers for this).

  28. CR
    ‘Taking seats off the SNP is neither here nor there, all that counts is taking seats from the conservatives’

    I understand that – though in terms of gaining an overall majority it will be important to win back seats from the SNP. As it happens, Labour also has realistic hopes of gaining 4 or 5 Tory seats in Scotland from third place.


    @”To my mind, despite the relative success of the GE for Corbyn and LAB, they need to move sharply towards the centre if they want to capitalise on the current situation.”

    No chance. And no sign of olive branches to Cooper et al.

    The Labour Party will be moved to a position where its institutions & governance ensure that the Far Left control it. Otherwise all of this euphoria ends with Corbyn, who may be the Messiah -but isn’t actually immortal here on Earth.

    So Parliament will take a back seat until they have reselected the “wrong ” MPS from Laverie’s “too wide ” coalition. Power will be sought on the Streets by the “People”. More & more demonstrations. And every time Cons do a u turn on policy they will say-see- more demos gets more change, and eventually a change of Government.

    The Canterbury constituency is the model I think they believe is there future.

  30. @Pete – “The public really doesn’t trust either of them to run the country.”

    When did the public last trust any party to run the country, though? Given that majorities in parliaments rarely reflect a majority in the nation, I’d say not in my lifetime.

    I’ll personally be letting the polls roll by for a few months before paying attention. I wonder when You Gov will start again?

    Agreed, and there may well not be many of them.

    Even if May is defenestrated PDQ and there’s a Con leadership contest over the summer we need to be a little patient, although it might provide some good clean fun on these threads.

    My guess would be that citizen’s rights, the exit fee and Ireland will keep Davis busy over the summer and nothing is going to be agreed before the German GE happens towards the end of September.

    I suspect we may have to wait a while for Scottish polling before having a feel for what’s happening.

    SLab’s leader helped SCon by announcing early on that she recommended voting SCon in seats where SLab couldn’t win. Whether she remains in post is another matter, but I don’t imagine that Lab HQ were thrilled by her intervention.

    At the same time, SCon haven’t exactly covered themselves with glory by NOT demanding Barnett consequentials for Scotland as a result of the DUP deal.

    Given that the Scottish media are strongly unionist, they may well hold off from polling for a while.


    @”they need to move sharply towards the centre if they want to capitalise on the current situation.”

    You’ve certainly provoked a lot of response to this one, and many different opinions. Here’s mine.

    My reading of both the Brexit vote and the GE is that there’s a growing number of people who are unhappy with the status quo, whether you the current policies left, right or centrist. Many sided with Labour this time as they were offering something different. Not enough to oust the Tories, and maybe that is to do with some of the leadership, but having a ‘different’ manifesto seems to me to be the crux.

    So depending how the current government respond, it may be that Labour moving to the centre ground will be perceived as ‘more of the same’. My feeling is that is exactly what they have to avoid, or they’ll be back in 2015 territory. But who knows really, there’s probably some time for re-alignment before the next GE.

  33. WES @ PETE
    Given that majorities in parliaments rarely reflect a majority in the nation, I’d say not in my lifetime.

    Nor mine, which began during the 1945 Parliament.

    Best scores of the big 2 since 1945 were:

    Con 49.7% 1955 Eden WON

    Lab 48.8% 1951 Attlee LOST

    If nothing else, it does demonstrate how little the plurality system represents the popular will.

  34. Colin

    The problem that Labour will have is if they go the union /demonstration route although it gives them air time on the BBC and Sky the general public will begin to tire of this tactic as it will begin to cause disruption in there lives like for example Southern rail.
    And It’s a perhaps sad fact of life that people although quite rightly outraged over things like the tower fire have soon grown weary of the continued political point scoring from Labour and from numerous commentators.
    I firmly believe although Corbyn has managed a surge in support on a personal level he will eventually come to grief at the hands of his own back benchers and the abrasive nature of his momentum supporters.

    As far as the Tories go like you I was very disappointed with the outcome of the GE as I believe although I understood the reasoning behind it it was badly run and couldn’t adapt as Labour brought forward it’s give away manifesto.
    However if they keep there nerve they command a majority in Westminster so it’s unlikely to be defeated in any policy decisions of course they will have to make concessions and if they have any sense they must raise taxes in the next budget so they can remove at least some of the public sector pay caps.
    The trick for them is to re establish there credibility as fiscally sound party with some small concession while making Labour look no more than a rather narrow protest party a difficult thing to achieve that will need a change after brexit of leadership.

  35. I listened to that podcast – thanks CMJ – and very interesting it was.
    IMO Lab is politically by and large where it needs to be (thank you JC – you have successfully moved the goalposts) but is in clear and present danger of self-immolating due (according to taste) to vengefulness amongst Corbynistas or undemocratic tendencies amongst the rest of the party (please don’t confuse that with ‘Blairites’)
    If JC decides to be inclusive and gathers all the talents, including those who have previously dissed him, Lab could become truly electable. If not, the Tories will grind him/them down and either there will be a change of leader or frittered support.
    He sacked the dissenting shadow ministers because he could, but actually it just underlines his insecurity.

  36. Hmmm. At a closer look on that Survation poll, it’s looking more and more bizarre. As a statistician I know of course that cross-breaks have a huge margin of error, in great part because samples are not necessarily representative within the cross-breaks themselves. But Survation’s really are all over the place in a way other polls just aren’t.

    As I said, in the latest Survation, Labour have 18% in Scotland, versus the Tories on 42%, SNP 29% and LD 10%. The margins of error on the subsample is obviously big, but Opinium’s manages to be quite close to the actual result in Scotland (it shows 38 SNP, 28 LAB, 27 CON). Similarly, in Wales, the Survation poll has the Tories on a 17 point lead, despite the lead being about 15 points in Labour’s favour in the election. Opinium has a 12 point lead for LAB in Wales, which is broadly right.

    The last telephone poll that Survation did had a 3 point lead for Labour, and showed an absurd lead of Lab over Con in their Scottish subsample (47-14 to Lab), though Wales was broadly in line with their GE2017 performance (44-30).

    With this in mind, if you look at the England figures, they show an interesting story. The margin of error on these figures is obviously much lower, but what they show is on the previous Survation poll, LAB and CON were tied at 45%, but even though overall the Tories lead by 1 on the new poll, in England LAB are 2 points ahead (44-42). For comparison, the Opinium poll has the England lead at 7 points (47-40).

    Ironically, I conclude this analysis of the two polls by suggesting we don’t look into them too much, seems to be largely statistical noise. (Though of course I don’t actually mean don’t analyse them, so much as don’t draw any serious conclusions from them)


    I think Dugdale is on borrowed time. Labour underperformed in Scotland in the election compared to many other areas which was inexcusable starting from such a low point. To recommend voting Con to beat the SNP was incredibly stupid and naive. She needs to go.

  38. GUY M

    My main observation on us as a people is that we rarely cling on to “heroes” for very long. Corbyn has a very limited window of opportunity – but also he can’t do a great deal about widening it as long as the present government continues in power.

    I am also not convinced that the Labour Party’s EU position is quite as clever as some think [inasmuch as I understand it at all].

    And I’m not really clear what it is that keeps the Tories polling around 40% in current circumstances; though I am clear that that is significant.


    As an FYI more than anything else: sacking ministers who defy a three line whip is common practice.

    As flagged by – I think – Andy Slaughter, not sacking rebelling ministers previously was the exception. Hence one minister resigning prior to the vote, as is good practice.

  40. Colin – all huff and puff and fewer than half a dozen Labour MPs will be deselected maybe only 2-3.

    Lavery’s piece is an invitation to former rebels to be patient and that included Cooper whose standing makes things trickier. I understand her and Starmer get on well and he in entrenched (Don’t want to say unassailable as we know where that end).

    Lavery wants a few rules changes endorsed or at least not opposed in particular the trigger for auto-reselection being higher.


    I’m no fan of Dugdale, but IIRC it’s up to SLab to choose their own leader. I can’t see her lasting long myself but they’re not exactly spoilt for choice amongst their MSPs and it makes more sense for their leader to be based in Holyrood rather than Westminster.

    That said, I’ll be mildly surprised if she’s still in place come the autumn.

  42. Just read this in a Guardian commentssetion:

    “Corbyn will get as far as Neil Kinnock did and that’s already happened..”

    I quite liked the irony but, from what I have been learning, in this very boutique, about ellipisis, I don’t think two dots is enough.

    I prefer loads of the little rascals myself………………………………….

    Yes, you’re quite right, of course. Just think that good leadership in Lab at the moment would look like compromise and the application of healing balm, rather than sacking people who by and large were taking a principled stance and setting off Lavery to talk of deselection.
    I think it’s part of a civil war, and I’m more or less a pacifist.
    A bit akin to Brexit, I believe if you take the view that those on the other side are b*stards who are out to get us, that is likely what you’ll get.
    Of course it’s true that the b*stards were out to get Corbyn before the GE, nor do they now think he’s the Messiah. But good leaders go out of their way to promote unity.

  44. Scottish subsamples from Opinium

    SNP 38%, Labour 28%, Conservatives 27%, Liberal Democrats 3%, UKIP 2%, Greens 2%.

    Labour, for all the faults of Dugdale, are back in front of the Conservatives and the SNP poll is steady.

  45. Scottish subsample from Survation has the Conservatives in the lead over the SNP – 42% against 29%

  46. Mike Pearce
    ‘I think Dugdale is on borrowed time. Labour underperformed in Scotland in the election compared to many other areas which was inexcusable starting from such a low point. To recommend voting Con to beat the SNP was incredibly stupid and naive. She needs to go.’

    I think that is being unfair on Kezia Dugdale. A mere 3 months ago it was being widely suggested that Labour would end up on 14%/15% in Scotland – whereas they ended up on 27% and just 1.5% behind the Tories with their representation at Westminster increased from 1 to 7. How many predicted that? This also followed outperformance by Labour at the May local elections. Kezia is the first leader in Scotland for many years to take Labour forward in seats and vote share. Moreover the party is now well placed to capture 20 SNP seats next time plus a few from the Tories and so go a long way to reverse the tsunami which swept it away in 2015. Kezia deserves some credit for that.

  47. Kezia going comments are being made by those who may not have an appreciation of the hole that Slab were in (and still are only a smaller hole).

    Unless she chooses to go she will lead Slab to the next Holyrood Election, which may well show big gains for the SNP vote share compared to the 2017 GE as split voting reasserts more prevalently; it would be wrong to blame her for this though it is just a fact of life in that polity.

    She may get to take SLab to the next GE as well when split voting should imo help them to second in seats terms and maybe sneak first?

  48. Graham

    It’s weird that the SNP are labour allies in Westminster but in Scotland labour is allied with the Tories. The 12 extra seats that the Tories won in scotland are the reason they are still in power. Of course SLAB expected the election to be a disaster for labour so taking bites out of the SNP and helping the Tories to do the same made tactical sense. But being wise after the event it looks an unfortunate strategy to say the least.

  49. In the survation poll 76% of Labour voters support ‘remain’. That can’t end well.

  50. Graham

    Sorry but I cannot agree. Labour still finished third in Scotland in a three horse race. Yes Labour went forward in vote share and number of seats but frankly it would have been difficult to do have done much worse. They were starting from an incredibly low base so the only way was up.
    Had Corbyn been more involved in Scotland during the campaign Labour would in my opinion have done considerabky better.

    Despite losing seats in England and Wales the Tories won a dozen in Scotland. Yes it’s partly down to the Unionist/ Indy debate but it’s also due to Davidson being an effective leader whereas SLab have the insipid Ms Dugdale.

1 2 3 6