Despair for Dugdale as SNP soar to 34% lead on the Holyrood list in tumultuous TNS poll

It's been a long time since we've had a full-scale Scottish Parliament poll.  The most recent one was the Ipsos-Mori phone poll which completed its fieldwork in mid-November, and showed a slight drop in the SNP lead.  Since then, one or two of our unionist friends (naming no names, but Aldo) have got carried away with the odd glimmer of hope in subsamples and a local council by-election in Blantyre (I know, I know), and convinced themselves that there are finally signs that Labour are closing the gap.  I fear that today's new TNS poll is going to be something of a hammerblow for them. 

Constituency ballot :

SNP 58% (n/c)
Labour 21% (-3)
Conservatives 12% (n/c)
Liberal Democrats 4% (n/c)

Regional list ballot :

SNP 54% (+2)
Labour 20% (-5)
Conservatives 12% (+1)
Greens 9% (+4)
Liberal Democrats 4% (-1)
UKIP 1% (-1)

A poll from TNS isn't the ideal way of breaking a long drought, because the firm's face-to-face fieldwork takes place over a period of weeks, and is always somewhat out-of-date by the time we see the numbers.  So there's still a theoretical possibility that there's been a very recent change in fortunes that this poll was unable to detect.  However, many of the interviews took place after the closure of the Forth Road Bridge (the latest in a long line of supposed turning-points for the unionist parties), and there's no sign of that having had any negative effect on the SNP's standing.  The Natalie McGarry controversy is also partly factored in.

As you may recall, the SNP scored 60% or higher on the constituency ballot in the first three monthly TNS polls after the general election.  They've been consistently below 60% since the late summer, so it looks like there was some genuine slippage after the post-May hoo-ha died down a little.  But it seems that the position has stabilised in recent months - the further drop to 56% in September now looks very much like a blip caused by normal sampling variation.  Weirdly, the SNP's 54% on the list ballot is a joint post-election high - it equals what they had when they were on 62% in the constituencies, and betters what they had when they were on 60% in the constituencies.  I can't think of any obvious explanation for that, unless SNP supporters are simply coming to the view that they don't want to split their two votes.  But, even now, almost half of the Greens' 9% support on the list is coming from people who plan to vote SNP on the constituency ballot.

There's no doubt that this poll will give the Greens a lot of heart after a string of disappointing findings for them (only Survation have offered them any comfort in recent months).  However, until their apparent bounce-back is confirmed by other polls, there remains the possibility that it's just an extreme example of margin-of-error noise.  And I'd certainly advise people to pay only limited heed to the excitement on Twitter about the Greens' 24% share of the list vote in Lothian, which is based on a regional subsample of just 85 people.

The biggest story of this poll is that Labour's mini-recovery since the spring seems to have been completely wiped out.  They were consistently on 23-25% of the list vote in the last three TNS polls, which was a few points higher than their showing in the early post-election polls.  But all of a sudden, they seem to be practically back to square one -  20% is just 1% higher than what they had in the May TNS poll.  Again, though, that may be a sampling blip - we'll just have to wait and see.

No such comfort for the Tories, who find themselves languishing on a dismal 12% of the constituency vote for a fifth consecutive month - that's 2-3% lower than they managed in the first two post-election TNS polls.  The pollsters are divided on whether or not the Tories are in a competitive race for second place, but if TNS are even vaguely close to being right, a few right-wing commentators are going to have egg on their faces after their recent musings about how their favourite party must be in line for a long-overdue breakthrough because Ruth Davidson is just so funny, so ballsy, and...ooooh, so smashing!

Irritatingly, TNS are still offering their respondents the SSP as an option, rather than RISE.  However, given that the SSP have once again scored a big fat zero on the list (or strictly speaking 0.2%), and given that RISE enjoy weaker brand awareness than the SSP, there is no particular reason to suspect that RISE would have registered any support in this poll.

There's more grim news for those who adhere to the Kenny Farquharson/Fraser Nelson worldview that Scottish public opinion is near-enough identical to English public opinion (once you strip away the inconvenient fact that the two countries keep voting for different parties).  One of the supplementary questions in the poll is about Britain's nuclear weapons, and the percentage of respondents who say that Trident should not be renewed significantly exceeds the percentage who say it should be.  (29% support renewal, 38% don't).  That's the opposite of what we know to be true about English public opinion on Trident, and it's a finding that should be taken very seriously, because this is not an online poll with a sample that is potentially skewed by having too many politically aware people - it's a 'real world' poll with a sample found by knocking on people's doors.