There's good news and bad news from the latest update of YouGov's projection model. At UK-wide level the model continues to defy the expectations of most commentators by pointing to a hung parliament, and indeed the Tories have edged down to their lowest number of projected seats so far (just 305, more than 20 seats short of a majority). However, there's troubling information from Scotland - it appears that the most recent batch of interviews must have been less favourable for the SNP, because their projected seat total has dropped from 47 to 42, which is their lowest so far. Even more worryingly, the SNP's floor (the minimum number of seats they'd be expected to win within a 95% confidence interval) has dropped sharply from 26 to just 17, which indicates that the Labour surge has reached the point where a large batch of central belt seats would be on the fringes of becoming competitive if YouGov are overestimating the SNP a little and underestimating Labour a little. On a more positive note, the SNP's ceiling is a much healthier 52 seats.
If there has been a recent dip in the SNP's support, I suspect their exclusion from Friday's Question Time special will have had a lot to do with it. Yes, that will be theoretically balanced out with a separate programme tonight featuring Nicola Sturgeon and Tim Farron, but that won't receive anything like the same publicity as Friday's show, which framed the election choice as being traditional and binary.
YouGov seats projection :
Conservatives 305
Labour 268
SNP 42
Liberal Democrats 13
Plaid Cymru 2
Greens 1
"Others" 1
Northern Ireland Parties 18
The most important thing is the SNP and Labour can work together to get the Tories out. Nothing is more important than getting them out of office.
ReplyDeleteWell, that all depends on whether YouGov are closer to the truth than ICM and ComRes. The nightmare scenario is if YouGov are correct that the SNP have slipped, and ICM are correct that the Tories are still headed for a three-figure majority across the UK.
DeleteThanks for ruining breakfast!
DeleteAny idea when the interviews were carried out?
ReplyDeleteWell, over the last seven days, but I'm not clear on whether they give more weight to the most recent interviews.
DeleteWell not long now folks. I think in 2 way fights in marginal seats between SNP and Labour. SNP would be too strong for them. Other than Murray's seat ,I am scratching my head trying to find a winnable Labour seat.
ReplyDeleteThere are moments like this when I wish that the SNP would have stood down and passed the baton to ConLab and said, OK, Scots, this is who you seem to want, let them try to governor FOR Scotland and not for London. Get back in touch when you've had enough"
ReplyDeleteThen the SNP would rightly be accused of bottling it, and abandoning their principles. Not a strategy worth pursuing, even if at times it seems plausible. No SNP=no interest or motive whatsoever for unionists in Scotland and the rest of the UK to even think about Scotland, let alone valuing it.
DeleteSo the ceiling is 52 - 17 now?
ReplyDeleteI.e. They don't have fkn clue?
There is a misconception that polling day will see vindication for certain polling companies and complete disaster for others.
DeleteIt's way too late for that.
Whatever the result political polling in the UK will take yet another battering where it matters. From the public's perception of it.
They public not give a flying fuck which polling company it was who was closest in the end. The public will only remember that in the run up to the election the one thing that was abundantly clear was that the polling was all over the place and therefore not worth relying on. Yet again.
Basically the ability of polls to shift public opinion has been proven to be utterly futile when it comes down to it. If that was the case then there would be zero ambiguity of the result from the polling weeks ago.
Nor did we need the EU referendum to already have proven that.
The public's perception of polling and where it was traditionally 'consumed' has been on a downward spiral since well before that. The growth of social media as a primary information source will not stop after polling day.
The other big factor is that terrorist outrages can't shift opinion much either politically or, again, the EU referendum result would never have been what it was.
There's a reason canvassing is now easily the most important and valued skill when it comes to informing a political parties approach to elections and the general strategy to take and implement.
If you don't have the infrastructure, activists or party resources to canvass heavily you are going to be whistling in the dark relying purely on completely contradictory polling sources who simply cannot be reconciled with each other.
Other problems aside, i think lots of people now intentionaly line to pollsters or use it as a way to influence things. I think people say they dont vote tory thinking it will get them to drop bad policies, say they are pro labor for same and want a threat. Also, the pollsters really suck.
DeleteI sense similar complacency to that of a few weeks ago when folk were denying the Tory rise in fortunes. It was real (if not huge) and they made council seat gains. Its the same story with Labour in Scotland now, its real (if not huge). Its up to the SNP campaign and activists to get out the SNP vote and not have any regrets on Friday morning.
ReplyDeletePaul I am not sure how you can say that, all SNP activists are out working hard and I can assure there is no complacency, if anything we are working harder than ever
DeleteI don't mean amongst activists (of which I am one too). I should have been more specific in suggesting some level of complacency in the Yes-voting 45% (for want of a better grouping), some of whom who have had difficulty getting their heads around firstly a Tory and now a Labour rise in the polls. We know the former will cost us a few seats (whatever us activists do).
DeleteThe projection is based on national share. In first past the post system national share is irrelevant. Hence the Libs can win a seat on 6% while Labour could win 17% and get no seats. It's an election in every single seat. For all we know Labour are taking percentages off the SNP in seats they can't win. So overall vote share goes down but not the seats we can win. We only got 49% of the vote share last time yet won 95% of the seats.
ReplyDelete"The projection is based on national share."
DeleteNo, it's not quite as simple as that - they do use constituency-level data wherever possible (and their projection for Orkney & Shetland is now looking a lot more plausible, probably for that very reason).
I can't recall ever seeing such a confused set of polls as this election. I would settle for 42 though. A clear win and potentially the balance of power. May would likely resign and Tory bloodletting ensue. Strange times
ReplyDeleteThanks James for clarifying.
ReplyDeleteHowever in terms of the national vote share, it's hard not to look at the headline numbers and read into it. It's a natural thing to be downhearted when you see a drop to 40%. However the difference between success and failure can be down to each seat even the knife edge ones. The big conurbations like Glasgow ,Lanarkshire and Edinburgh are where the biggest percentage of national vote share are gathered.
We know the SNP are unlikely to lose any Glasgow seats even if the margin of victory is lower. So losing 3% nationally gathered in Glasgow overall would result in no change to seats. The smaller constituencies in terms of numbers like Ochil and Moray don't have as big an influence on overall vote share.
What I am trying to say is that if people switch back to Labour in Glasgow from SNP. It will result in a decline in national vote share but have no impact on the seats in Glasgow.
For what its worth, 42 seats is my personal prediction. The SNP may not have slipped, or at least not by very much. The model may now have had enough interviews to factor in tactical voting in specific seats: for example I see that the 4 Lib Dem targets have now gone to them.
ReplyDeleteIm persoanlly concerned that according to the model, Mhairi Black is in great danger - Labour seem very close in Paisley and Renfrewshire South. However they may be overstated since the Labour vote was artificially high here in 2015 due to a high profile incumbent and heavy targeting.
I think there must have been some slippage, or at least a net swing of some description from SNP to Labour - I can't see how else to explain the SNP's floor slipping as low as 17. If memory serves me right, the floor was still in the low 30s only three days ago.
DeleteMhairi Black has a majority of 5000 - 12%?
DeleteJames what would that mean for the likes of Glasgow, Paisley, etc
DeleteYouGov have the SNP ahead in all the Glasgow seats as far as I can see, but Labour are too close for comfort in Glasgow South-West (although of course the model is untested, and individual constituency projections have to be treated with extreme caution).
DeleteThanks James. So we need to be very cautious of the model then and its projections?
DeleteJames, the SNP have a majority of nearly 10,000 in Glasgow South West. Can we really treat their projection in that seat as being reliable at all?
DeleteI've already said that individual projections should be treated with caution, but no, it's not inherently implausible that the SNP could lose a seat like Glasgow South-West on a bad night. If the vote comes down quite a bit, that's going to vary from place to place - in some seats, the drop is bound to be higher than average.
DeleteElectoral calculus has the SNP on 48 seats?
ReplyDeleteElectoral Calculus is rubbish. (That's a technical term.)
Deleteso which is the most reliable in your opinion James?
ReplyDeleteThe last u gov poll I did and one's before tend to lump the Scottish Nationals in with the Welsh nationalists
ReplyDeleteHow they can predict seats from that is beyond me
Most of the postal votes will already be in. At the council elections they accounted for a lot of the turnout. I have a feeling they likely favour The Mooth. So the volatility is likely to be amplified in the 50 to 60 % of the vote still to be cast. I don't think you can call this election at all.
ReplyDeleteCan we ask Ruth the Mooth how many PV votes have been cast and for whom? I did read somewhere that 27% of the votes cast at the Cooncil Elections were postal votes. The Tories did very well in those elections.
ReplyDeleteHow many deid folk have voted? Well the SNP have done nothing to reform the PV system as far as I can see. So if they are reduced to less than 50 seats then it will be a case of we told you so and the SNP did nothing to address a voting and electoral system that Mugabe would be proud to have.
The SNP have no control whatsoever over voting arrangements for Westminster elections.
DeleteHow accurate/ useful is Tactical2017. ?
DeleteIt has some severe limitations. If a non-Tory candidate won the seat last time, it just automatically tells you to vote for that candidate's party, regardless of whether tactical voting is actually required in the constituency. For example, people are told to vote Liberal Democrat in Orkney & Shetland because that's a Lib Dem-held seat, which is completely silly because there's no danger of the Tories winning there.
DeleteThis is sounding grim .The party are going to regret diluting the Indy message for this election.I am now resigned to losing a dozen seats or even more .
ReplyDeleteThe SNP are being non-stop slandered by Scottish libdems, tories, labour AND vast majority of the media. With all this having a majority is nothing short of a miracle. I honestly don't think diluting the indy message has much effect. What we're seeing is Unionists voting Tory, and the SNP losing some votes to Labour because of Corbyn. Unfortunately Scottish labour and UK labour are different beasts, so voting for Scottish labour is not voting for Corbyn, but for red tories.
DeleteWe won 56 seats on the back of the 45 despite the media .Labour have shook up the election and inspired the young .Whilst the SNP have ran a boring defensive campaign.I get that campaigning hard on Indy was a risk but being defensive is going to costrich us a dozen seats or more .I am just a bit worried we will be seen as losing momentum
DeleteThe westminster bubble media and BritNat spin on the result is immaterial.
DeleteThey were always going to attempt to portray an overwhelming win for the SNP in scotland as a crushing defeat.
Just like they tried and failed to do at the council elecs.
What matters is we win a convincing majority of MPs to carry on the mandate we already have forward as the BritNat unionist parties stagger witlessly into the oncoming Brexit catastrophe.
The Labour party is still dominated by MPs who despise Corbyn and will never stop undermining him until he is gone. Most notably in scotland where Dugdale and her SLAB chums are still demanding Labour supporters vote tory to 'stop' the SNP and the next Indyref.
The tories are still led by an obviously incompetent leader who hasn't the faintest idea how to handle Brexit or the breathtakingly inept far-right fools she nominally put 'in charge' of the process.
If the BritNat yoons think that is a recipe for success in the weeks, months and years ahead then they have truly taken leave of their senses.
@Annoymous
DeleteNicola Sturgeon has done what she needed to do earlier this year, namely signalled an intent for a second independence referendum, before Article whatever it was, was triggered. That is now done, and Holyrood has voted in support of another referendum as well. There was always going to be a reaction from the more hardline unionists in Scotland to this. We just have to keep slogging away at the moment, to see what the political climate will be like after Thursday's vote, and then the resulting EU negotiations. My guess is many unionists know that they are going to turn into a disaster. If we have to suffer some short term pain for long term gain then so be it imo.
I generally hate the Guardian but - I can barely get the words out - it seems to have just backed the snp.
ReplyDeleteNot sure that'll sway many voters but quite interesting
It is of no consequence.
DeleteThe Grauniad has long ceased to be relevant to grassroots
and left-wing Labour, never mind relevant to the inner political workings of a faraway country of which they no little.
It will at least anger the Dugdale Hothersall SLAB BritNats and there is certainly much amusement to be had from that.
However, SLAB are fast becoming as relevant to scottish politics as the lib dems. Even stealing back a poor second place from the tories would hardly convince anyone other than themselves that they are actually for anything. They are merely Red Tories and Ruth Davidson's unionist wannabes.
As long as SLAB put British Nationalism before everything else they will be seen as a pale imitation of SCON and pay for it at every single election.
Curtice has just tweeted a TNS poll. As usual, it's a bit dated already (because it's collected by face-to-face interviews).
ReplyDeleteSNP 43
Con 28
Lab 21
LD 5
https://twitter.com/WhatScotsThink/status/871797094662766593
And here is where the glib analysis on thus posting falls short., there is a vote shift in O&S -Tories gaining Lib Dem and SNP Leave vote. Still not clear what it means in the end but not a foregone conclusion. Opinion polls are pulse-taking. Of course they can change. Events dear boy, events.
ReplyDelete