Full-scale Scottish polls from Survation are like the proverbial London buses - you wait a whole month for one, and then two come along on the same night. But who's complaining? The second one is even better than the first...
Scottish voting intentions for the May 2015 UK general election (Survation, 12th - 17th March) :
SNP 47.1% (n/c)
Labour 26.1% (-1.9)
Conservatives 15.5% (+1.0)
UKIP 4.3% (+0.9)
Liberal Democrats 4.0% (n/c)
Greens 2.2% (-0.1)
The percentage changes listed above are from the first Survation poll of the evening, which is the correct approach because that's the last directly comparable poll. The Daily Record (and most other sources) are reporting different changes, because they're measuring from the last Survation poll commissioned by the Record itself. On that basis, the SNP are up 2%, and Labour are down 2%.
I've already had to correct this post, because I originally stated that this was the poll we had the sneak preview of yesterday, but in fact the fieldwork dates don't match up with this one either. I'm guessing that the SNP's question about keeping the Tories out must have been a bolt-on to part of the fieldwork for both of tonight's polls, which effectively leaves us with a phantom 'composite' voting intention poll, albeit one that isn't filtered for turnout.
The independence question has also been asked...
Should Scotland be an independent country?
Yes 50.9% (+3.2)
No 49.1% (-3.2)
This is the first time ever that a Survation poll has shown majority support for independence - which is pretty extraordinary when you consider that Survation were consistently one of the most Yes-friendly pollsters until the "Great Convergence" at the very end of the referendum campaign. The highest Yes vote they showed in a pre-referendum poll was 48%, and since September they've produced two narrow No leads and one 50-50 tie. They were much quicker than YouGov to introduce weighting by recalled referendum vote, which means that this is only the second time that any firm has shown Yes ahead in a poll using the new weighting (the previous example was a Panelbase poll in the autumn).
People who voted Yes in September have been downweighted in tonight's poll from 442 to 414, and people who voted No have been upweighted from 496 to 511. By definition, that procedure wouldn't have happened in pre-referendum polls. If Survation had retained their old methodology, Yes would now be ahead by quite a bit, and would certainly be doing much better than the pre-referendum record high of 48%. There can be no doubt at all that support for independence has increased markedly since polling day - all firms that have asked the question are agreed on that point.
As in previous Survation polls, respondents were asked when there should be a second independence referendum. Only 20.1% replied that it should never take place. A total of 40.3% want it at some point within the next five years, and a total of 59.4% want it at some point within the next ten years. So nothing very different from the consistent picture we've been seeing over recent months. The people of Scotland seem to have reached a clear judgement that they want a say on independence again before the semi-mythical "generation" is up, and they certainly want it well before David Cameron's breathtakingly arrogant timescale of "a lifetime" has passed.
Last but not least, here are the Scottish Parliament voting intention numbers. As before, the percentage change figures are from the first poll of the evening.
Scottish Parliament constituency ballot :
SNP 49.8% (+0.6)
Labour 26.3% (-0.6)
Conservatives 13.3% (+0.4)
Liberal Democrats 5.0% (n/c)
UKIP 2.5% (+0.3)
Greens 2.0% (-1.1)
Scottish Parliament regional list ballot :
SNP 39.4% (-1.9)
Labour 22.8% (+0.3)
Conservatives 14.6% (+2.7)
Greens 11.1% (-1.5)
UKIP 5.5% (+0.5)
Liberal Democrats 5.3% (-0.4)