Westminster in crisis tonight as new Norstat poll shows support for independence has risen to around 50%

One of the minor wins for the unionist camp over the last couple of years is that they’ve somehow persuaded media organisations and some polling companies to revert to publishing independence poll numbers with Don’t Knows left in. That shouldn’t be happening – the practice was completely abandoned in the run-up to the 2014 indyref, and for obvious reasons you would never see voting intention numbers for Holyrood or Westminster that don’t exclude undecideds. The purpose of the wheeze is presumably to give the false impression that support for independence is lower than it is – if you say, for example, that “Yes support is at 47%”, most people will assume that means No support is at 53%, whereas 53% is in fact the combined figure for No and Don’t Know.

In the case of the latest Norstat poll for the Sunday Times, the figures with Don’t Knows stripped out haven’t even been mentioned in the write-up at all, so all I can tell you is that a rough recalculation shows that the headline numbers must be either Yes 49%, No 51% or Yes 50%, No 50%, hence my use of the phrase “around 50%” in the title of this post. It’s extremely frustrating not to know for sure, but the fact that we don’t can be assumed to be an intentional tactic. The data tables will resolve the mystery in the fullness of time.

Should Scotland be an independent country? (Norstat/Sunday Times, before exclusion of Don’t Knows)

Yes 47% (+2)

No 48% (-1)

This is the first poll since Norstat took over Panelbase, so the percentage changes are measured from the last Panelbase poll. If anything, Panelbase have tended to be one of the more No-friendly firms in recent years, which makes Yes practically drawing level all the more striking as a development. From a personal point of view, given my history of commissioning polls from Panelbase for Scot Goes Pop, I’m sad to see that the Panelbase branding seems to have been dropped for this poll, and I’ll be interested to discover whether the change is purely presentational or goes deeper than that.

While this is an unalloyed good news poll for the Yes movement, pretty much the opposite must be said for the SNP, who in Westminster terms have suffered their worst showing in a Panelbase/Norstat poll for many years.

Scottish voting intentions for the next UK general election:

Labour 36%

SNP 33%

Conservatives 16%

Liberal Democrats 7%

Seats projection (with changes from 2019 general election): Labour 28 (+27), SNP 18 (-30), Conservatives 6 (-), Liberal Democrats 5 (+1)

Paradoxically, there’s some comfort to be drawn from the fact that Humza Yousaf’s net trust rating in the poll is -25, and Nicola Sturgeon’s is almost as bad at -19, which would have been unthinkable until very recently. That suggests there is a personal element to the SNP’s current unpopularity, which at least in theory could be reversed with a change of leader. Whereas Tory rebels look a bit odd plotting to bring down Rishi Sunak when there is no obvious election-winning replacement waiting in the wings, the SNP are fortunate enough to have a ‘queen over the water’ figure in the shape of Kate Forbes, who we know is more highly regarded with the public than Yousaf.

Scottish Parliament constituency ballot:

SNP 36% (+1)

Labour 31% (-)

Conservatives 16% (-)

Liberal Democrats 7% (-)

Greens 5% (-2)

Scottish Parliament regional list ballot:

SNP 30% (+1)

Labour 29% (-1)

Conservatives 19% (+1)

Greens 9% (-3)

Liberal Democrats 7% (-1)

More details and analysis to follow shortly…