Labour now trail the Tories by 25% in England (and by a lot more in the over 75s)

I’m putting this here for those who wish to comment, from the ICM poll for the Guardian. First thing to note is that there is no statistically significant change, but the position is not great for Labour…

Conservatives: 43% (up 2)

Labour: 26% (down 2)

Ukip: 11% (down 2)

Lib Dems: 8% (down 1)

Greens: 6% (up 2)

Now, 17% is likely to be accumulated damage from a summer of very nasty and public in-fighting. You pays your money and you takes your choice as to who is to blame for that. But the poll suggests that other critical gaps are even wider…

And although in the case of 18-24s Labour has a solid advantage, amongst the very oldest (75+) the slip in pro-Labour sentiment is shocking (just 7%)… And whilst just a year ago, Labour was deemed to have been struggling to get 36.7% in England…

…the Conservatives now lead in England by 25%. Labour are reckoned to need to get their noses ahead by 13% in England to beat the Tories.

The Referendum seems to have been the turning point. Labour is struggling (as they did after the Scottish poll) to account for why critical numbers of their own constituents backed nationalist right arguments rather than a left-led Labour party...


  • chrisjones2

    Just like Burnham did when he was Health Secretary?

  • mickfealty

    They’re going to need an awful lot more than that… The gender gap is particularly deadly. But so too is England as an entity…

  • hgreen

    Good job it doesn’t matter until 2020 by which time Scotland will be off and May will have had to press the UK self destruct button.

  • mickfealty

    It won’t be that easy Hugh. This is rapidly becoming my favourite Scottish political site. Scotland’s dilemma will not be so simple to resolve as you seem to believe (…

    Nicola Sturgeon’s decision on the timing of the independence referendum is likely to be the most important of her leadership. If she gets it right her place in Scottish history is assured but if she gets it wrong her time as First Minster would be over. It can be very lonely at the top.

    For one, timing is not in her gift, but in her main rival Theresa May. That will constitute a sort of preliminary that may be hard to trigger (and win) before the negotiation of Article 50 is complete and made public.

    If the economic weather is sh!t, May will resort to the SNP’s own once in a lifetime line to continue to refuse. She’ll dig in hard, and argue the SNP now has a wheen of powers it asked for which it should get on with using.

    Meanwhile, Boundary changes, and Labour will have Jeremy, Diane or John at the dispatch box, and whole bunch of newly selected candidates going for far fewer viable seats in 2020.

    (And remember, it’s not just the ones they are going to lose that matters to a Labour, it’s the changes to the ones they had a reasonable chance of getting back that will also be gone from view).

    The leadership is locked in till impact. Labour MP’s will do what Paul Ryan is now telling his senators and congressmen and women to do, which is tell themselves “feck the general election and hold on to what you can”.

  • Declan Doyle

    You are bang on the money there regarding Nicola, but i wouldnt underestimate her ability to spot exactly the right moment to go for it.

  • mickfealty

    Nor me in fact.

  • Obelisk

    Nicola will only go for the referendum when she knows she can win it.

    A position I wish some of my fellow Nationalists would adopt in regards to the border here.

  • chrisjones2

    The logical thing for Theresa would be to announce that if the Scottish Assembly demand independence by a vote within the next 3 months she will immediately set the process in train with an exit by 2019.

    If they vote for it it will be economic suicide and financial oblivion. They will procrastinate. They might even be forced to call an election to get a mandate

    Labour will be all over the place as they see all those seats potentially disappearing

    Then sit back and watch the SNP squirm and prevaricate to avoid the fateful day when yet again the electorate decide that they cannot afford independence

  • Megatron

    Statistically significant means nothing as a term.

    Please stop using it. Why not just say changes are small or something?

  • Declan Doyle

    Here its different. The dynamic and circumstances are different. We are seeing a steady rise in support for unity. If SF suddenly stopped calling for a border poll they would be accused of sell out etc. It would also fuel a growing agitated dissident threat and potentially fracture the party. If support for unity can be maintained and built upon and if the new boundaries manage to deliver electoral parity the circumstances will evolve to justify a poll.

  • Obelisk

    I am not against agitating for unity by any means. I believe it is the best outcome for the North which faces being left behind by the rest of the world.
    Nor do I begrudge Sinn Fein calling for one because I understand that may be for strategic reasons….showing they have not given up.

    I am just opposed to the sort of person who, when they would have looked at those polls showing low support for unification would not only call for a border poll but actively insist on it.

    I believe that a border poll should only be called when we stand a good chance of winning one. And the reason for that is that a loss would be hurled in our faces by Unionists endlessly after.

  • Declan Doyle

    But there is another important element to that. Calling a poll will focus minds and weed out the intent of southern parties in particular. From SF point of view, for so long as a poll looks like a far out or unlikely prospect, those parties can sit back and relax without any inquisitive research or proposals. And thats the key. No one party can deliver unity, a united nationalist push on an agreed platform is far more likely to succeed.

    30% are in favour of unity with no debate, no programme, no sign of a poll, no initiative from other nationalist parties, no support from Dublin and no accomodation from London. Thats an incredibly solid base and way ahead of where Scotland started just a few years ago.

  • hgreen

    Heard all that nonsense about financial oblivion the last time. Why as a britexshiter is it ok to say we want our country back despite the negative economic consequences and not OK for the Scots to do the same?

  • Reader

    Megatron: Statistically significant means nothing as a term.

  • OneNI

    ‘we are seeing a steady rise in support for unity’ LOL
    Support for unity in elections and polls is failing.SF and SDLP both LOST seats in last Assembly elections.
    I assume your mention of new boundaries refers to Westminister – think ‘parity’ unlikely
    Also Conservative and Unionist Party likely to have thumping majority influence of Irish nationalist handful (half of whom obligingly make the current Conservative Govts life easier by not turning up) will be neglibile

  • terence patrick hewett

    By which time there will be more over 75’s

  • chrisjones2

    Yes it does – its very precise …or imprecise perhaps … the article

  • Megatron

    Ok I should have been a bit more precise myself. Statistically significant only means anything when it is framed by a hypothesis test and some definition as to what your confidence level should be. Given
    that was not set out above it means nothing.

    I will concede that If the hypothesis is that there has been no change in party support, and you have set your confidence level at 95%, the statistical results here not “significant” enough to call your hypothesis into question. Why we would be interested in hypothesis testing at 95% level is beyond me and maybe someone can enlighten me. Why not set it at 90% or 99%?

    The main problem is that saying the changes are not statistically significant glosses over the fact is that with a 2% change there is a high probability that support has changed.

  • Kevin Breslin

    No more UKIP to split the Tory vote …

  • john millar

    “Why we would be interested in hypothesis testing at 95% level is beyond me and maybe someone can enlighten me. Why not set it at 90% or 99%?”

  • chrisjones2

    “the fact is that with a 2% change there is a high probability that support has changed.”

    No it doesn’t. It all depends on the sample size. For example if change is +/-2% and that is inside the error limits for the poll it means nothing at all. It could simply be a random error / quirk

    Your hypothesis here is that a chance of 2% represents a change in overall opinion

  • Megatron

    Eh not sure the link answers the question? I think I just about passed statistics 101 thanks very much. Any idea why we are so focused on the 95% confidence level?

  • chrisjones2

    Ah “britexshiter” – you have gone back to your usual position of personal abuse because you cant cope with the arguments and facts.

    The Nats financial plan for independence was based on oil at $100 . Its now $48 leaving a revenue hole in the hypothetical Scottish budget big enough to swallow the entire Scottish NHS and a chunk of pensions post independence

    So if the Scottish people want independence and realise they cannot then afford services on the current level, yes that is their choice. I know what way I think they will vote and I suspect Wee Nicola knows that too which is why she has been a little subdued on the subject

  • Megatron

    “No it doesn’t. It all depends on the sample size. For example if change is +/-2% and that is inside the error limits for the poll it means nothing at all. It could simply be a random error / quirk”

    That is Exhibit A why we should not use the term statistical significance. The statement you made is completely incorrect.

  • chrisjones2

    a united nationalist push

    Combined Loyalsist Military Command

    UUP / DUP Pact

    All equally effective

  • chrisjones2

    We are slowly taking over

  • woodkerne

    Arguably, we are already living in a one-party state. 26% was about the same level Labour achieved in the 1983 election when last led from the left of the party’s traditional coalition of interests. Tony Benn rationalised it at the time as 6 million socialist votes! And this may indeed represent the level of real support in the electorate at large for a Corbynist (aka retro Bennite) left-Labour policy platform. In a PR election, allied pre- or post-campaign, formally or informally, with the Greens and Welsh, Scottish and northern Irish nationalists, a progressive-bloc of this order (accumulatively approaching circa 45%) would surely be enough to block the right (tories and kippers) from power. But the 2020 election will not be fought on PR and even with some manner of pre-electoral pact, the Labour Party as things stand are set to be annihilated. Indeed, unless the left in UK politics wakes up soon and commits to the realpolitik of anti-tory coalition building, the prospect is of a further decade or more of right-wing government. Of course all kinds catastrophic possibilities lie just ahead. Interestingly, in policy terms, a sound bet is that demand-led investment in infrastructure plus a positive role for the interventionist state are positions likely to be embraced across the parliamentary spectrum. One feasible scenario is that corresponding splits on the right and left in response to post-brexit tensions could occur. So although it is too soon in the cycle to predict Labour wipe out, the indications aren’t good and Labour isn’t currently in a fit state to capitalise on the inevitable fall out post-Article 50.

  • john millar

    Its our old friend the standard deviation of the Gaussian “bell curve”
    In any population ( People /things etc) there is an average (mean) of some kind (height /age weight )

    Based on the bell curve there as many below an average as there are above it

    68% of values are within
    +- 1 standard deviation of the mean

    95% of values are within
    +- 2 standard deviations of the mean

    99.7% of values are within
    +-3 standard deviations of the mean

    at 95% you can claim to represent “most” of the data via an “average”

  • Megatron

    at 95% you can claim to represent “most” of the data via an “average”
    Surely at 68% you are representing “most” of the data? So why not use that value?
    3rd time lucky…why are we focusing on 95% confidence interval?

  • Declan Doyle

    You are wrong, the most recent lucid talk poll shows support for unity at 30%, up from just 7% a year earlier.

    The loss of one or two seats is hardly a crushing blow. Remember also that PBP took two seats.

    The bew constituencies are far more evenly balanced and offer a potential boost for nationalism. Check Faha’s analysis over on bangordub.

  • hgreen

    Personal abuse and then you use the sexist phrase “Wee Nicola”? Really?

    Again my point is, should you care to address it ,why is it ok for Brexshiters to ask for their country back despite the negative economic consequences and not the Scots?

  • Lex.Butler

    30%, after nearly 100 years of partition and numerous attempts at a violent reunification, doesn’t strike me as a solid base but the opposite. Since every ‘push’ to date has failed, what’s your magic formula that will suddenly convince the Shankill and Ballymena (for starters) to embrace nationalism? Or do you just ignore them (again)?

  • Declan Doyle

    Thirty percent is where scotland was two years before their poll and that conversation was in full flight.

    Nobody should be or needs to be ignored, everybody who wants to engage can, whether they are ballymena, Shankhill, the waterside or Portadown.

    The people will decide by way of a democratic vote. Its pretty simple.

  • mickfealty

    Simon Nixon in the WSJ (

    An Ipsos Mori poll of 1,013 people in Northern Ireland between mid-August and early September found that only 22% of voters currently support reunification, albeit up from 17% in 2013. Indeed, only a minority of voters even want a border poll, with 52% opposed.

    The only way of increasing that figure to the required level is incrementally and over time by building faith and credibility and seeking to converge the material interests of both parts of Ireland.

    Oh, and not pretending to be stupid or ignorant of the reasons why partition happened in the first place (more a feature in southern politics than in NI I might add).

  • Zig70

    I don’t care that Corbyn isn’t popular. I think it will take a bit of time to relearn vocational politics. Once labour plp realize it is about what you stand for and not how you can spin yourself into power to maintain a salary and benefits that many politicians act like they are entitled to.

  • john millar

    “3rd time lucky…why are we focusing on 95% confidence interval?”

    in statistics, we are usually presented with a sample from which we wish to estimate (generalize to) a population,

    How much confidence in an estimate do you want?

    Calculate the average of a sample of data from a population Now calculate the standard. deviation. of the same data

    Then for 67% confidence the average of the WHOLE POPULATION will lie in the range sample average + or – 1 sd

    For 95% confidence the average of the WHOLE POPULATION will lie in the range Sample average + or- 2 sd

    How much confidence do you want?

    Maths really is fun

  • Megatron

    Ok this might be shorter if I point out I have a degree in maths and statistics and don’t need educating on the basics of statistics. My job is statistical analysis. Math is indeed fun.

    Now for the fourth time…why are we picking 95% as confidence interval instead of 67%, 50% or 99%?

    You seem to be asking the question back to me for some reason. I am asking why people who have chosen 95% have chosen that number.

  • Declan Doyle

    The lucid poll was the most recent and the most under reported. In any event there’s a bit of momentum there without natuionalism lifting a finger so lets see how it progresses.

    Your second point is bang on. It will be slow and hopefully steady.

    Your last comment is a bit of a hit and run; care to expand for clarity?

  • OneNI

    ‘Remember also that PBP took two seats’ Yes precisely people voted for Socialists over Nationalist/Republicans

  • john millar

    “why people who have chosen 95% have chosen that number.”

    For a Gaussian distribution, almost all data will fall within three standard deviations of the mean. Broken down accurately the empirical rule shows that 68.27 % will fall within the first standard deviation, 94.45% within the first two standard deviations, and 99.73% will fall within the first three standard deviations of the distribution’s average.

    When presenting data to the innumerate it is common to approximate these figures to 68% and 95%.

    This allows opinions to be expressed

    “68% of the data lies between X and Y “ (mean +- 1s.d.)

    95% of the data lies between A and B “ (mean +-2 s.d)

    Most presentations I have been involved in use both percentages — because of the perceived greater” confidence” implied by the higher percentages

    (95 is more than 68)

    I presume that is why 95% is used in this case

    ( You could ask the company which carried out the survey)

  • chrisjones2

    Heightist not sexist

    There you go again with Brexshiters. You just cant help yourself or are you just trolling?

    Sometimes things that need to be done have a price (short term) for a benefit (long term). You are just appalled that the majority of our voters voted to exit. I am sorry for that but hell, thats democarcy

  • mickfealty

    Cannot find the LucidTalk poll. Have you a link? Meanwhile, as Peter noted previously, whatever it is the NI nationalist parties have been doing, it’s not working:

  • lizmcneill

    Rising prices of groceries will also hit the elderly on fixed incomes hard.

  • Declan Doyle

    Turnout is shocking low. But Brexit, the growth in popularity of PBP and the possibility of FF getting on the horse just might shake it up.

  • MainlandUlsterman

    I see Owen Jones is once again daring to question the cult of Corbyn … the running dog