Scottish Parliament constituency voting intentions (Scot Goes Pop / Survation poll, 11th - 13th January 2021):
SNP 51% (-2)
Conservatives 19% (-1)
Labour 19% (-1)
Liberal Democrats 9% (+3)
Scottish Parliament regional list voting intentions:
SNP 40% (-1)
Labour 19% (-1)
Conservatives 17% (-1)
Greens 11% (+1)
Liberal Democrats 8% (+1)
Seats projection (with changes from the 2016 election): SNP 69 (+6), Labour 23 (-1), Conservatives 19 (-12), Greens 11 (+5), Liberal Democrats 7 (+2)
This is the third Survation poll in a row to put the Tories in third place on the list - and the one prior to that had them in joint second place with Labour. However, there's no sign of Labour making even the kind of modest recovery hinted at by the Westminster numbers from this poll - they're actually slightly down on both Holyrood ballots, and are facing a small reduction in their number of seats. They're only in second place by default because the Tories are doing so badly.
Another key difference with the Westminster results is that the SNP vote isn't down by as much. A two point drop on the constituency ballot, and a one point drop on the list, could well just be margin of error 'noise'.
65 seats is of course the target for an overall majority, so the SNP's projected 69 seats would put them well over the top. And the SNP and Greens in combination are projected to have 80 seats - meaning that pro-independence parties would have an extraordinary 62% of seats in the parliament.
Meanwhile, tonight has brought word of the twentieth poll in a row to show a pro-independence majority (or nineteenth in a row if you exclude one that asked a non-standard question).
Should Scotland be an independent country? (Panelbase)
Yes 52% (-4)
No 48% (+4)
The percentage changes are from the most recent Panelbase poll, which was the one I commissioned in November, and which put Yes on an all-time high for Panelbase of 56%. The four-point drop for Yes needs to be seen in that context - after a record-breaking result it's not unexpected to see a reversion to the mean in the next poll. That said, the last four Panelbase polls have all shown Yes on between 54% and 56%, so tonight's 52% may indicate a genuine reduction (with one possible explanation being the Sturgeon/Salmond controversy). On the other hand, if the true Yes vote had remained static on around 54%, the occasional result two points higher or lower than that would be totally normal due to the margin of error. As ever, we'll just have to await more information from more polls.
UPDATE: Someone on the previous thread suggested that Panelbase may have asked a non-standard question this time, which would obviously render the percentage changes meaningless. But I'm not sure whether that's really what happened, or whether the Sunday Times gave a misleading impression with their summary of the results.