Alba vote increases in new Savanta ComRes poll

I didn't in any way blame Savanta ComRes for the #Matchettgate episode, because I'm quite sure the fault lay at Scotland on Sunday's end.  However, I think eyebrows are going to be quite rightly raised at the ComRes spokesman's comments about today's new poll, because they do appear to dispense with political impartiality in two key respects.  It's a poll that shows pro-independence parties with a very clear majority of seats in the next Scottish Parliament, and yet the ComRes spokesman feeds into a Tory narrative by casting doubt over whether that would represent an "unequivocal mandate" for a referendum.  I'm not sure it's for a pollster to suggest that a majority of seats isn't a proper mandate in a parliamentary system.  

He also describes the Alba Party as "ever-unpopular", which is a truly extraordinary remark.  Alba are a minority party in the hunt for seats in a proportional representation system - just like the Greens, just like the Liberal Democrats, just like the SSP in the old days.  Nobody would call the Greens "unpopular" just because 90% of people aren't going to vote for them.  The ComRes poll actually shows the Alba vote increasing slightly, which reinforces the message from other recent polls that their support is holding up.

Scottish Parliament constituency ballot (Savanta ComRes/The Scotsman):

SNP 45% (-1)

Conservatives 23% (-2)

Labour 23% (+3)

Liberal Democrats 7% (+1)

Scottish Parliament regional list ballot:

SNP 36% (-2)

Conservatives 22% (-1)

Labour 19% (+2)

Greens 10% (+3)

Liberal Democrats 5% (-)

Alba 2% (+1)

Seats projection (with changes from the 2016 election): SNP 61 (-2), Conservatives 28 (-3), Labour 24 (-), Greens 11 (+5), Liberal Democrats 5 (-)

Pro-independence parties: 72 seats (55.8%)

Anti-independence parties: 57 seats (44.2%)

PRO-INDEPENDENCE MAJORITY OF 15 SEATS

There are quite a number of plus points here - although Savanta ComRes have joined Panelbase in showing the lowest SNP constituency share since 2019, the SNP lead has actually increased due to the Tories dropping back.  And in the constituency section, it's the gap between the first-placed and second-placed party that really matters.  The 25% Tory share in the last ComRes poll always looked a bit odd in the context of other polls.   And it now seems pretty likely that the relatively poor Green showing in the last ComRes poll was a freakish finding caused by sampling variation - 10% in the new poll is much more in line with others firms' polls, and it naturally has a very positive effect on pro-indy representation.

Should Scotland be an independent country?

Yes 46% (-2)

No 54% (+2)

(UPDATE: I've corrected the percentage changes above because I was accidentally looking at the last-but-one ComRes poll as the baseline.  The actual numbers in the last poll were Yes 48%, No 52%, so the swing this time is 2% rather than 4%.  Bear that in mind when reading the analysis below.)

I don't think there's much doubt that Yes support has dipped a little over the course of the campaign, which is what traditionally happens due to the SNP campaigning on bread and butter issues and leaving the pitch clear for unionist parties to relentlessly bash independence.  But there is *considerable* doubt over whether Yes have dipped by anything like as much as ComRes are reporting.  At some point soon I'll be releasing the independence numbers from the Scot Goes Pop / Panelbase poll, which had overlapping fieldwork dates with ComRes.  Part of the reason I held the numbers back for a few days was just in this case this sort of scenario occurred.

I've had two analysis pieces about the Scot Goes Pop poll in The National over the last couple of days - you can read them HERE and HERE.