Scot Goes Pop / Panelbase poll on independence: Yes storms back into the lead in wake of Dominic Cummings episode

For the second time this year, Scot Goes Pop asked for your help to commission an opinion poll on independence, just to check whether a landmark recent event had boosted support for Yes.  The first time around in January, we got the result we wanted - Yes had jumped into the lead with 52% of the vote.  I did worry that it was too much to hope that lightning would strike twice, but what do you know?  It has.

Should Scotland be an independent country? (Scot Goes Pop / Panelbase poll, 1st-5th June 2020):

Yes 52% (+2)

No 48% (-2)

As always, the headline results exclude Don't Knows.  With Don't Knows left in, the results are Yes 48% (+2), No 45% (-1).  Percentage changes are measured from the last Panelbase poll, which was conducted on behalf of Wings Over Scotland early last month.

This is the fourth Panelbase poll on independence in 2020 so far.  The sequence of results for Yes has been 52 - 49 - 50 - 52.  All of those results are theoretically within the margin of error, so there are a couple of ways of interpreting what we've seen - it could be that the Yes vote has been holding steady over the last few months after an initial post-election surge, and that the differing results have been caused by random sampling variation.  Or it could be that the surge subsided just slightly in late winter and early spring as attention turned away from Brexit and towards the coronavirus crisis, and that there has now been a bounce back for Yes as a result of the Dominic Cummings episode and the UK government's mishandling of the pandemic.  I must say that looking at the results of the supplementary questions from this poll (which I'll release over the coming days), I'm inclined much more towards the latter theory.

The datasets show a now-familiar pattern.  A large minority (35%) of people who voted Labour in the general election would vote Yes.  Significantly more No voters from the first indyref (18%) have changed sides than Yes voters (8%).  That's different from polls a few years ago that showed post-2014 progress for Yes without any net movement at all from No voters - back then it was 2014 non-voters who were making the difference.  That suggests to me that the current Yes lead is built on much more solid foundations.

Naturally, most Yes voters are Remainers.  56% of people who voted to stay in the EU would now back independence.  But in fairness a healthy minority of Leave voters (34%) are on the Yes side too.

The results of the poll were, in line with standard practice, weighted by both 2019 Westminster vote and 2014 indyref vote.  The latter led to Yes voters being downweighted somewhat, although nowhere near as drastically as is sometimes the case.  (Among those likely to vote in a new referendum, 387 Yes voters were reduced to 371 in the weighted sample.)  It's worth making the point that 2014 is now six years ago, and there may come a point in the coming years where pollsters will conclude that weighting by the indyref result is no longer sensible, simply due to the sheer passage of time and the increasing danger of false recall.  If and when that happens, it will probably result in at least a slight average boost for the reported Yes vote share.

There are several more questions to come from this poll - I'll be releasing the details bit by bit.  If you'd like to be the first to know every step of the way, you can follow me on Twitter HERE.