Sunday, December 14, 2014

Misery for Murphy as sparkling SNP seize sizeable Scottish Parliament lead in new YouGov poll

I've only got a couple of minutes to spare before I pop out for the rest of the day, but I thought you might be interested to see the Holyrood figures that have now been released from the new YouGov poll...

Holyrood constituency voting intentions :

SNP 50% (+4)
Labour 28% (n/c)
Conservatives 14% (+2)
UKIP 3% (-1)
Liberal Democrats 3% (-2)
Greens 2% (-3)

Holyrood regional list voting intentions :

SNP 42% (+4)
Labour 26% (n/c)
Conservatives 14% (+2)
Greens 7% (-3)
UKIP 4% (-2)
SSP 3% (n/c)
Liberal Democrats 3% (-1)

Even with the Greens slipping back somewhat, it remains the case that the SNP's hopes are threatened most by people who voted for them on the constituency ballot drifting off to the Greens or SSP on the list ballot.  That's fair enough if those people actually regard the Greens or SSP as their number one choice, but if they instead think they're doing it as some kind of pro-independence "tactical vote", it's folly in the extreme.  As the 2011 result amply demonstrated, "tactical voting" on the list is a mug's game, and has at least a 50/50 chance of either not working or backfiring completely.

On a more positive note, it's worth remembering that it was on the list vote that the pollsters were miles out in 2011, so it could be that an aggressive "vote twice for an SNP government" strategy will do the trick again.

It doesn't look like YouGov have weighted by recalled referendum vote, because excluding didn't votes/can't remembers/refusers, the recalled Yes vote is 47.1%, compared to an actual Yes vote in September of 44.7%.  That's a pretty narrow gap, though, so the impact would be relatively modest even if an additional new weighting was introduced.  On the question asking for voting intentions for a hypothetical new referendum, Yes are slightly ahead even on the raw unweighted data.  Compare that to the famous YouGov poll showing Yes ahead ten days before polling, when No remained ahead on the unweighted numbers.  There really does seem to have been a sea-change over the last three months.

UPDATE : A point that's just occurred to me is that it would actually be wrong for YouGov to weight by recalled referendum vote, because they're only using over-18s for their current polls, and nobody has a clue what the referendum result was if 16 and 17 years olds are excluded.


  1. Shame the SNP increase is coming from Greens and Libs rather than Labour. But at least the polls aren't showing an increase in the Labour vote either.

    There's no way Jim Murphy can turn this round by May. There are far too many folks like me who voted Labour in 2010 to keep the tories out but who are utterly sickened by them after the indyref and who have since joined the SNP.

    You'll never win us back, Jim, but do keep trying, it's fun watching you squirm!

  2. Jim Murphy claiming Labour won't lose any Scottish seats to the SNP just shows how little they understands about what is happening in Scotland.

    It's as hilarious as Ruth Davidson claiming the Scottish Tories will increase their vote to 50% in the GE. That's how out of touch Labour are, they haven't a clue.

    Jim Murphy would cry if he came to a gathering of my ex-labour-voting family who now hate their guts.

  3. Disappointing to see the Greens aren't pushing ahead on the list. The only way to kill off NuLabour is for the Greens to take away their list seats. Currently an SNP list vote is a wasted vote in the vast bulk of Scotland.

    1. The list vote was not wasted in 2011. That is how the SNP achieved its overall majority. The SNP won every constituency in the North East and still gained a list seat. In other areas it won list seats in addition to the FPTP seats won

    2. That's a scary misconception. It's the list that determines the balance of the parliament, and if the SNP only get constituency seats, we're screwed.

    3. Even in 2011 when SNP support was below its current Constituency level, a list vote for the SNP was completely wasted in Lothian and in two other regions only returned one MSP for the highest list vote. I see the indications that the SNP Constituency return is likely to be much higher this time round.

      I am not an SNP member or activist, I am a supporter of an Independent Scotland. The most important thing *to me* is wiping out the Unionist voices at Holyrood and 50% seat SNP and high list Green would move towards this.

    4. OK, I see your point, I'll think about it.

  4. "Blow for Sturgeon as SNP unable to make impact on Labour vote."

    "Labour vote holds up under new Murphy leadership"

  5. I see in the Sunday Times poll the SNP/Plaid sample was 92 but weighted down to 37.

    1. Highest down-weighting since the referendum (and that's accounting for overall sample down-weighing to match population share) and, historically, at the very high end of SNP down-weighting. Serious SNP suppression anyway.

      As I mentioned in a past post, down-weighting of SNP has been growing since the iref steadily. At around 0.85x now (taking into account population share adjustment as noted) on average.

      Kellner correction is still in use in Scottish Yougovs. That has always hurt the SNP and no reason to suggest it still isn't.

      Could mean SNP is a good bit higher than Yougov are getting. Not implausible given MORI numbers (no past vote weighting). The next MORI will be one to watch.

  6. If the SNP are polling 45+ in the run-up to the election, I'm sure there'll be a lot of campaigning by pro-independence groups to educate the population that a regional vote for the SNP is nigh on useless.

    If most of the SNP support voted Green, SSP or Solidarity with their Regional vote, Labour could lose half their MSPs.

    On current polling the LibDems would be down to around 2 MSPs

    1. Dangerous to assume the SNP won't need list seats to form a government. Anything could happen.

    2. Not really. I will vote SNP, but if they need the Greens to keep them in power I won't be crying in my soup!

      Let's take the North East example mentioned earlier...
      140k votes equaled 1 list MSP for the SNP

      If say 50k of them had gone Green they could have gained 2 seats, and say another 20k votes went to the SSP they'd gain a seat, at the expense of 1 Tory, 1 Lab and 1 SNP.

      Thats with only half the SNP vote shifting. Is 2 Green and 1 SSP MSP fair swap for 1 SNP?
      Thats open for debate.

    3. I think you'd need a crystal ball, but we'll see. Depends on how the polls are looing by May, too.

    4. Boab : The example of the northeast demonstrates exactly why "tactical voting" on the list is a crazy idea. I lost count of the number of people in 2011 who said "I'm voting tactically for the Greens on the northeast list because the SNP can't win a list seat here" - and in the end the SNP took a seat on the northeast list while the Greens didn't. Anyone who did vote "tactically" put that SNP seat at risk without having any real prospect of getting the Greens over the line.

      There are just too many variables for tactical voting on the list to be realistic - you're relying on a precise degree of accuracy in opinion polls that just isn't available.

    5. Yes there is always a speculative choice in voting, even without lists not making or making a tactical choice requires speculation. But we will have a very strong indication in Westminster 2015, if the SNP are successful in taking 35 seats or more, the idea that they will get more than 60 FPTP Seats at Holyrood 2016 becomes compelling.

      Once that is the case, the number of list seats in *some* if not all regions starts to approach zero, effectively leaving any List vote for the SNP as worthless as those in Lothian in 2011 where 110,000 list voters chose the SNP and had a return of zero.

    6. "the number of list seats in *some* if not all regions starts to approach zero"

      "All"? If they were winning 60 constituency seats Scotland-wide, why would they do worse on the northeast list than in 2011, when they won a list seat in spite of sweeping the board in the region's constituencies?

    7. Perhaps I'm misconstruing what you said - the asterisks might be intended to indicate that you accept it won't happen in all regions.

    8. In the North-East the SNP only just won a seat with 140k votes and didn't get any in Lothian with 110k. If the polls are showing that the SNP are likely to sweep the board, tactical voting could be hugely advantageous to pro-independence parties such as Greens and SSP, if well rallied, and organised.

      Like I said, I'd rather have 3 pro-independence non-SNP than 1 SNP. That's just me.

      But then, we really are getting WAY ahead of ourselves - its a long way off yet!

    9. "if the polls are showing that the SNP are likely to sweept the board"

      But they're not going to show any such thing - except in the same sense they "showed" that a tactical vote was a good idea in the northeast in 2011. What can possibly go wrong? Well, you only need to be 100% sure that the SNP will do well enough (sweep the board in the constituencies), but not too well (have enough support to nick a seat on the list anyway). You also need to be sure that the Greens aren't being overstated, and that you aren't throwing your vote away on a party that has no chance of winning a seat in the region.

      Opinion polls of that precision haven't been invented yet.

    10. But we have an actual election 10 months before Holyrood 2016. Westminster will tell us what is happening in terms of FPTP results and how much the SNP can still gain there. A quick scan of the Holyrood 2011 results shows that the majority of FPTP seats the SNP do not hold are held by their incumbent by less than 1500 votes and within easy reach of the relatively small increase in SNP and decrease in Labour currently shown in polling.

      The only Region where there is a definite prospect of SNP gaining list seats outside of South (although all the seats have the SNP close it was obviously the least pro-Yes region) and North due to the Lib Dem strength in the Orkney and Shetland Constituencies (why aren't they merged, i mean seriously they are tiny). The probability of persuading enough support to get North East style results elsewhere even in the North East again is pretty slim as far as I can see.

  7. It’s often hard to follow Murphy’s reasoning. His claim that Labour won’t lose any seats to the SNP in May looks like a reckless hostage to fortune. Murphy is not exactly stupid. He’s a good enough tactician when it comes to climbing the greasy pole or manipulating the mainstream media (not that many of them need much manipulation). But where’s the strategic thinking? What’s the political principle? How does he plan to win back the leftist voters who have defected to the SNP, the Greens or the SSP?

    Maybe he has been persuaded by his own success at constituency level, holding on to an East Renfewshire seat that is often described as natural Tory territory. Maybe he sees a reinvigorated Scottish Labour as the flag-bearer of a Better Together alliance, harvesting Tory, Lib Dem and even UKIP votes to thwart the SNP. It’s interesting that the Telegraph is interpreting his assertion that he will make his own policies in Scotland as a Blairite putsch aimed at Miliband. Maybe he does think Scottish Labour can win by moving to the right.

    This may be Machiavellian but it is not totally dotty. We know from preference voting in recent local by-elections that many Tories prefer Labour to the SNP. But how will traditional Labour voters react if Murphy starts cosying up to the Tories? Will their new hatred of the SNP outweigh their ancestral hatred of the Tories – especially when English Labour will be concentrating their fire on the self-same Tories?

    The structural consequences, if by any chance this scenario ever happened, would be fascinating: Labour would become an avowedly left-wing party in England and an avowedly right-wing party in Scotland. Maybe Murphy is a closet nationalist after all.

    1. Thinking about it from a purely electoral strategy perspective, I'm not convinced moving to the left actually would be the best strategy for Labour. If we were to put Labour to the right of the SNP (even marginally) you've got three parties at least to your left (SNP, Greens, SSP) who have some aspirations of entering Holyrood in 2016. If you're Labour you can't beat the Greens on environmental politics, you can't beat the SSP on left-wing politics and you can't beat the SNP among independence supporters. The Lib Dems will probably sack Clegg after 2015 so under a new leadership it's quite possible they'll be competing on the left as well - I certainly can't see them going to the right given how much of a beating they've taken for being associated with the Tories.

      In contrast on the right of the party spectrum you have a political pariah in the shape of UKIP, while the Tories still have a major problem attracting support in Scotland even among people who agree with them on policy (simply because their "brand" is so politically toxic). Competing for the centre ground to the right of the SNP is probably the best strategy for Labour in all honesty. If you're a No voter and don't want the Tories/UKIP there isn't really much of an option beyond Labour so I imagine Murphy's strategy will largely consist of simply trying to appear credible and not as bad as the Tories.

    2. If Labour are walloped in May, with Murphy saying they won't be, what will they do? Sack him, after 5 months? Seems doubtful. They have nobody else, as shown by his election.
      The only way Jim will go is if he loses his own seat or Labour loses the GE and he sees an opportunity down south.

  8. When was the fieldwork done for this one? Despite the amusing headline, I assume it concluded before Murph's anointment yesterday?

    1. 9-11th Dec.

      Murphy's been Branch Office Leader in waiting for ages now. BBC has made sure everyone knew that. The contest was just something that had to be done to keep up appearances.

      Did anyone seriously believe it was going to be Findlay or something?

      The recent boost for the SNP could well be the Murphy effect in action.

    2. "Murphy's been Branch Office Leader in waiting for ages now. BBC has made sure everyone knew that."

      I'm not buying that I have to say. Everyone expected David Miliband to win in 2010 and he lost to his brother in much the same kind of scenario (Ed being the unions' candidate, David being the more polished of the two). I find it very hard to believe even if it wasn't a done deal that people in an opinion poll were actively "punishing" Murphy before he'd even been elected leader. In fact that's pretty ridiculous to put it mildly.

    3. Sorry, I should have added 'mild sarcasm' tags to my post.

      Murphy won't make a blind bit of difference to SNP VI.

      Personally, I never for a second thought it would be anyone other than Murphy. Labour do not learn from mistakes. Arrogance is too great. They probably don't believe polling at the moment. It just can't happen you know - Scotland not voting Labour at Westminster. It can't ok. It just CAN'T!! AAAAAARRRGHH.


    4. I agree, Harrylaird. There was enough speculation about the result that many spectators must have thought there was at least a chance of Findlay winning. If nothing else, SLAB will expect some kind of bounce in the next poll just from the media attention over the weekend.

    5. @Harrylaird

      Google each canditate bbc 2014 as the last time I did it gave JM 520ths NF 69ths and SB 46ths showing who Boothman and co wanted elected.

    6. "Murphy won't make a blind bit of difference to SNP VI."

      I'm inclined to agree and I thought Murphy was the best man for the job. I doubt there will be any Murphy bounce and I would imagine that if the polls do tighten, it would happen with or without Murphy.

  9. Lib Dems in 7th place in the regional vote! Well, maybe 6th equal. They'll soon be getting relegated!

  10. I've had to delete a comment from Anon because it contained a claim that could potentially be construed as libellous. For the record, Sean Thomas was charged with rape a number of years ago but was found not guilty as his trial.

    I'd just ask people to be very careful about making untrue claims, even when it's only intended as banter.

    Here is the deleted comment, slightly edited -

    "The Greens deserve to be stuffed for their abuse of the electoral system.

    Mike (Scots are subhuman filth) Smithson insists that labour in Scotland will hold all of their seats as the english dominated betting market is in their favour. Man is both deranged and a bigot.

    Not as bad as the hideous sean thomas. Apparently labour will win the 2016 election because Devo-Max will give the Scot's Parliament so much power that the electorate will be scared of an SNP Government.

    PS. Miss Plato. A prime example of the unthinking english racist. Truly vile."

    And Tris left this short reply -

    "It would be scary, right enough, if they changed ALL the road signs..."

    1. I'm not sure it adds much to the sum of human knowledge even in its redacted form.

  11. @Stoat "I doubt there will be any Murphy bounce"

    I imagine there will be a bounce simply due to huge publicity he is getting.
    The BBC news page is giving him free reign at the moment, with several high profile stories in a row, unchallenged by any opposing voices.
    Perhaps this is a standard 'welcome' for a new leader, I don't know..

    The current SNP will probably fall naturally as the GE approaches, as the focus on London parties intensifies. No doubt the media will try to spin any drop in the polls as the start of the Jim Murphy revival.

    He does have weaknesses though.
    All the pro-Scotland platitudes and soundbites we are seeing are empty rhetoric only.

    People know that Labour are weak on devolution, and would swiftly backtrack on even the feeble Smith proposals without any pressure from the SNP.

  12. It doesn't look like YouGov have weighted by recalled referendum vote, because excluding didn't votes/can't remembers/refusers, the recalled Yes vote is 47.1%, compared to an actual Yes vote in September of 44.7%. That's a pretty narrow gap, though, so the impact would be relatively modest even if an additional new weighting was introduced.

    As you say, as with YouGov's last poll at the end of October there is a slight bias towards Yes voters - 47.1% said they voted Yes rather than the actual 44.7%. This means that although it (again) looks as if Scotland would vote Yes by 52%, once you correct for the sample it reduces this to 49.1%. So it would still be No

    It shows where the movement is though - 97% of Yes voters would still vote Yes while only 85% of No voters are still No. The remainder have moved to Yes or uncertainty.

    Roger Mexico

  13. "So it would still be No"

    It might be either Yes or No, with or without any additional weighting - it's well within the margin of error either way.

    1. I think even at its most conservative estimate of 49.1% that's still very significant three months after the referendum. I'd be happy to accept we've seen a 4.4% bounce for independence so close after the vote, and I certainly don't see support for it going anywhere but up in the next few years. My prediction is we see an average in the polls by summer/autumn 2015 showing support for independence solidly within the mid-50s.

    2. What I would really like to know is a more accurate quantification of the Quebec Effect - that of people intending to vote Yes who switch when faced with the reality of the ballot slip. The best we seem to have is guesses that it's somewhere between 1% and 5% but this is such a wide margin that is becomes problematic.

  14. Populus sub-sample has SNP 36, Labour 28 and Tories 24. Fieldwork done between Friday and Sunday, i.e. it's the first work done (partially) covering the period after Murphy's election.

    Populus typically show a small SNP lead or sometimes a small Labour lead. Today's figure is a bigger SNP lead than the figures Mike Smithson was getting worked up about last week.

    1. Populus is fairly useless for polling at the moment because of its weighting. Their method of weighting by current Party ID to historic targets[1] means that they stabilise VI at that level and dampen down any subsequent movement considerably.

      With UKIP this meant that they kept on showing them at around 8%, even when every other pollster had them (at least) consistently over 10%. Then they adjusted their targets at the start of this February and UKIP leapt to being always at 13-14% (at a time when no other pollster showed any such movement). Since then UKIP has continued at that level, even though other polls showed further rises. You can see this pattern if you flip between the pollsters on the NS's Drilldown app.

      Because the Greens and SNP have risen in VI since the last 'revaluation' the movement shown is minimised and their support understated. If you look at the weighting in Populus's most recent poll, while Labour and Conservative are left pretty untouched, the SNP and the Green have their 'core' vote weighted down to a half and UKIP to one third. They then seem to add a fiddle factor to bring UKIP up quite a bit (I can't get the figures to match otherwise) but the SNP and Greens are left untouched.

      One particularly stupid thing about this method is that the more the allegiances of recent converts firm up and they become more likely to identify with their new Party, the more their vote will get down-weighted. So the more successful and established a Party becomes, the lower its VI with Populus will go. It also means that it's possible for panel members to make their reply count for more by lying about which Party they identify with, though I haven't seen much evidence of this.

      This is different from weighting historic allegiance to historic targets as YouGov do, which is a good way of getting around the problem of potential bias in online panels. To be fair to Populus they didn't have that option as they only started their panel after they last general election, though I suspect now they wish they had asked people when they joined, rather than each time.

    2. Roger, to return to the earlier point, it's only just occurred to me - it would actually be wrong for YouGov to weight by recalled referendum vote, because their current polls only include respondents who are over the age of 18, and by definition no-one knows for sure what the referendum result was among over-18s.

  15. 81% of 2010 Lib Dem voters are planning to abandon them in May. Ouch.