YouGov poll shows support for holding an independence referendum has soared

So this is a sort of request post - a couple of people asked me to write about a detail in last week's full-scale Scottish poll from YouGov that perhaps didn't receive due attention.

In principle, do you think there should or should not be a referendum on Scottish independence at some point in the next five years?

Should be a referendum: 45% (+3)
Should not be a referendum: 44% (-4)

The choice of question may seem a tad odd given that the Scottish government are proposing to hold a referendum a lot earlier than five years from now, but the wording was used to maintain consistency with the question that's been asked for a couple of years.  That means we can make a direct comparison with previous results, and I've commented a number of times before on the odd results this question has tended to produced.  Even when Panelbase were suggesting the public were split right down the middle on whether there should be a referendum in as little as two years, the YouGov question was stubbornly producing a solid majority against a referendum within the next five years.  It wasn't immediately clear why that was happening, as the YouGov question isn't in any way leading, so the diverging results could only have been a 'house effect' caused by the composition of YouGov's panel, or by their sampling, or by their weightings.

But whatever the reason, the fact that there is now a slim pro-referendum majority (once Don't Knows are excluded) must be seen as highly significant.  According to the What Scotland Thinks archives, this is the sixth time the question has been asked since April 2017, and on four of the five previous occasions 51% or more of respondents were opposed to a referendum.  The narrowing of the anti-referendum lead to just six points in the last YouGov poll looked dramatic enough, but now that it's been wiped out completely, the question arises as to whether other polling firms that have previously shown an even division in public opinion would show a decisive pro-referendum majority if they released a poll now.  It doesn't necessarily work that way, but it's a logical possibility.

Of course the main independence question in the new YouGov poll showed a no change position - it was 49% Yes in the spring, and it's 49% Yes now.  I saw a few silly suggestions from unionist commentators (taking their cue from Willie Rennie) that this was a sign that Scottish voters are shying away from independence due to the current demonstration of the chaos caused by a constitutional upheaval.  The reality is that as the Brexit crisis deepened earlier this year, the Yes vote in YouGov polling jumped to an unusually high 49% - in recent years the normal range in YouGov polls has been between 43% and 45%.  And that 49% has been maintained in the new poll - the new converts to Yes don't seem to be developing cold feet.  The most that can be said is that Scots perhaps didn't find Brexit under Theresa May any more palatable a prospect than Brexit under Boris Johnson.  But the changes on the 'do you want a referendum?' question suggest that there may indeed have been post-Boris movements in public opinion beneath the surface that haven't fed through to the main independence question yet.  Sometimes supplementary questions do give you a better guide than voting intention questions (for example leadership ratings are sometimes better predictors of election results than standard party political polling).

Regardless of the majority in favour of a referendum, it's still not clear how a referendum will actually come about.  Sometimes it's easy to lose sight of the obvious, so I don't think we should totally exclude the possibility that the SNP will secure the balance of power at the forthcoming general election, and will be able to win the concession of a Section 30 order as part of a deal to install a Labour-led government.  In the past, journalists have tended to assume that the SNP would have no real leverage in that scenario because they'd know they would pay too high a penalty for doing anything that might return the Tories to power.  But the electoral threat from Scottish Labour may now have receded to the point where the SNP won't feel they have much to lose from playing hardball with Labour in post-election negotiations.  And I'd suggest any future Section 30 order should permanently transfer the power to hold a referendum, rather than just for a time-limited period.

*  *  *

John Bercow's last stand against "not a usual" prorogation a couple of hours ago is surely destined to become the stuff of political folklore, but we're also seemingly heading towards something else that is highly unusual: a general election in November.  Since 1979, the practice has always been to hold elections somewhere between April and June, presumably to take advantage of longer days and better weather.  Snow isn't totally unheard of in November, and given this country's inability to cope with unusual weather, that could cause chaos.  For example, if an independence referendum had been held on St Andrew's Day 2010 as the SNP government had originally wished, it would have taken place on a day of heavy snow and severe traffic disruption.  The credibility of the result would probably have been called into question.

*  *  *

As expected, Stuart Campbell has topped off several days of abusive behaviour directed at this blog by blocking me on Twitter - which means I am now automatically on the notorious 'block-list' that he tries to persuade all his followers to use.  So please be aware of that if you're one of my followers on Twitter and if you wish to continue following me - using the block-list will lead to you blocking me without realising it (along with, I believe, another couple of thousand accounts, including some very surprising names that no indy supporter would want to block without good reason).