Apologies for the delay, which was partly due to the Mail doing the tedious (and fruitless) #buyapaper routine, and partly due to me having a nap. Here are the numbers from today's bumper Survation poll -
Scottish Parliament constituency vote :
SNP 56% (+2)
Labour 20% (-4)
Conservatives 14% (+1)
Liberal Democrats 7% (+2)
Scottish Parliament regional list vote :
SNP 45% (+1)
Labour 19% (-2)
Conservatives 12% (+1)
Greens 11% (+1)
Liberal Democrats 8% (+2)
UKIP 5% (n/c)
The percentage changes are from the last Holyrood poll conducted by Survation - but that was before the general election, so the methodology has changed since then (to introduce weighting by recalled 2015 vote). Although on the face of it the increase in the SNP vote is within the margin of error and therefore not statistically significant, the last poll was unusually good for the SNP as well, so there is very little room for doubt that the position has strengthened over a period of months. The only question is whether it has strengthened still further since May. When you take this poll in combination with the extraordinary TNS poll a few weeks ago, the most likely answer is 'yes'.
Again, we have the familiar pattern of the SNP vote dropping off sharply on the all-important regional list vote. That wouldn't harm the party on the current figures, but of course the likelihood is that they're in a honeymoon spell at the moment, and that support will slip back by next year. If, by then, the SNP vote on the two ballots is not 56% and 45%, but 45% and 36%, Nicola Sturgeon will fall well short of an overall majority. So unfortunately, what's happening on the list is potentially a very real problem.
Survation have tended to be one of the more Green-friendly pollsters. The 11% Green vote should therefore be treated with caution - if it's accurate, it would certainly point to the possibility of a meaningful breakthrough in terms of seats, but on the other hand it offers no support for the far-fetched claims that Patrick Harvie could somehow become leader of the official opposition.
Liberal Democrats 7%
As we now have weighting by 2015 vote, these numbers imply real changes from the outcome in May. A 1% increase for the SNP is not statistically significant, but a further 3% drop for Labour looks somewhat more interesting. Unless Survation have come up with a freakishly anti-Labour sample, Kezia Dugdale clearly has it all do to get her party back into the game, in both parliaments.
Should Scotland be an independent country?
This is very much within the range of results that Survation have produced since the referendum - mostly there has been a very slender No lead, but there has also been one Yes lead and one tie. The fluctuation could well just be margin of error noise, and we've seen much the same thing from other polling firms. The most plausible interpretation is that there was a modest jump in the Yes vote immediately after the referendum (probably caused by buyer's remorse among a small number of No voters after they realised The Vow was a con), and that the position has remained reasonably stable since then.
As always, it's important to stress that we can be confident that the Yes vote is genuinely higher than it was on polling day, because weighting by recalled referendum vote is now in place. Almost 5% of No voters from September have switched to Yes, compared to 2.4% of Yes voters who have switched to No. There are also significantly more people who have switched from No to Don't Know.
If there was to be another referendum on Scottish independence when, if at all, do you think this referendum should take place?
Within 10 years?
This is the answer that respondents have been consistently giving over recent months. They oppose a referendum within the next five years by a 59-41 margin, so clearly the centre of gravity is that Indyref 2 should take place at some point between 2020 and 2025.
Should the United Kingdom remain a member of the European Union? (respondents in Scotland only) :
An impressive margin for Yes given that this was an online poll (the online method tends to favour No), but the fieldwork took place between the 3rd and 7th of July - well before the "waterboarding" of Greece.