You might remember that at the weekend, I mentioned that Scottish data from the YouGov projection model had been revealed on Twitter, and it showed voting intentions of : SNP 42%, Conservatives 26%, Labour 25%. The reaction of our resident Tory troll Aldo was "Not a poll. Nothing to see here." Well, here's the bad news, Aldo - a full-scale Scottish poll was released by YouGov today, and it shows almost identical figures.
Scottish voting intentions for the UK general election (YouGov) :
SNP 41% (-1)
Conservatives 26% (-3)
Labour 25% (+6)
Liberal Democrats 6% (n/c)
Greens 1% (-1)
Regardless of whether the projection model is technically "not a poll", the GB-wide voting intention numbers it's been producing are strikingly similar to YouGov's regular polling, so it's no great surprise to discover that the same is true in Scotland. Nevertheless, it's still very reassuring to see more evidence that the Scottish Tories haven't been immune to the slump suffered by their colleagues in England and Wales.
On the face of it, this poll suggests that the SNP's support has been extraordinarily stable throughout the campaign. The first YouGov poll after Theresa May called the election put the SNP on 41%, the second put them on 42%, and today's puts them back to 41%. That looks very much like trivial fluctuation caused by margin of error noise. However, we shouldn't forget that both Panelbase and Survation have reported a modest SNP dip as Labour have recovered, so it's possible that the margin of error is actually masking a similar SNP dip in the YouGov poll. Labour have a slight lead among 18-24 year olds, and it's hard to believe a surge like that could happen without harming the SNP's overall vote at least a bit (unless it's been offset by the SNP regaining some votes direct from the Tories). But you'd be a brave person to look at the evidence we have so far and conclude that Labour's recovery has harmed the SNP more than the Tories. On balance, it looks like the opposite may be true - today's 15% gap between SNP and Tory is the joint-biggest that any of the familiar online pollsters have reported during the campaign in a conventional poll.
The first question that formed in my mind when I saw today's poll was "what were the fieldwork dates?" Yesterday's update of the projection model gave the distinct impression that SNP support may have slipped very recently, which obviously meant that a poll with older fieldwork wouldn't offer much reassurance. It turns out that it was conducted between Thursday and yesterday, so quite a bit of it was before the Corbyn v May Question Time special, and also before the tragedy in London. However, the fears raised by the projection model have eased for the time being anyway, because the central figures in today's update show the SNP bouncing back from 42 to 46 seats. Perhaps even more significantly, the SNP's floor (the minimum number of seats they'd be expected to win within a 95% confidence interval) has jumped back up from 17 to 32. That implies the gap between SNP and Labour in the popular vote has widened - although whether that's because of strengthening SNP support or Labour slippage is hard to say (short of obsessively checking the projected percentages from all of the individual constituencies, which I haven't been doing).
There doesn't appear to have been Labour slippage south of the border, though. The updated projection shows the lowest number of Tory seats to date, and suggests for the first time that the SNP and Labour in combination would outnumber the Tories in a hung parliament.
UK-wide seats projection (YouGov) :
Liberal Democrats 12
Plaid Cymru 2
Northern Ireland Parties 18
HUNG PARLIAMENT : Conservatives short by 22, Labour short by 60
On those figures (which may well be in the realms of fantasy, but who can say for sure?) it's unlikely that either a Tory or Labour government would be viable in the long-term - instead there would effectively be a caretaker government until a new election could be called, probably in the autumn. But it's interesting to speculate who would lead that caretaker government. Strictly speaking, it really ought to be Jeremy Corbyn - by constitutional convention, the Queen is supposed to appoint a Prime Minister who can command a majority in the House, and if the SNP, Plaid Cymru, Greens and SDLP all clearly signalled that they preferred Corbyn to May, it would leave Corbyn with the stronger case. But you could be sure that the Tories would be claiming that the largest single party should automatically get the nod if an opposition alliance can't muster an absolute majority, and you might well see right-wing Labour MPs sabotaging Corbyn by publicly agreeing with that line of argument (echoing the extraordinary behaviour of Tom Harris immediately after the 2010 election).