Scottish polling seems to have settled into a roughly monthly rhythm at the moment, though after months of Panelbase polls it’s nice to have a bit of a change with a YouGov. This one was carried out from the 6th to the 10th of August. Alongside and as a consequence of the ongoing Coronavirus Pandemic, this came in the week the Scottish Government were under heavy scrutiny concerning the estimation of SQA results when no exams had taken place. That was an ideal time to poll, to see if the issue was having an impact.
Please note this was quite a comprehensive poll, so this coverage is lengthy!
- Party – Vote% (Change vs last poll by agency / Change vs last election or referendum)
Holyrood Voting Intention and Projection (Tracker)
- SNP – 47% (+2 / +5)
- Conservative – 21% (-2 / -2)
- Labour – 14% (+2 / -5)
- Liberal Democrat – 6% (-1 / +1)
- Green – 6% (-2 / -1)
- SSP – 4% (nc / +3)
- Brexit – 1% (+1 / +1)
The answer to the question of impact was, apparently, “hee haw”. The SNP have continued their upwards growth, this being their best list result figure with YouGov since 2016. Almost everyone else is down versus both the April poll and 2016 election. Of the Holyrood 5 only Labour have noticeably improved by either measure, gaining a couple of percent since April, but leaving them with a figure that would still be their worst-ever at Holyrood. Perhaps to balance out an unusually favourable figure with Panelbase last month, the Greens are on an unusually unfavourable result for YouGov, this being the (joint) lowest the pollster have found them since 2016.
Note again that inexplicably large SSP figure. Figures of 3% date back as far as October 2017, so, I’m still very much of the view that this is a specific oddity of YouGov rather than any reflection of the SSP doing their best Lazarus impression. Nonetheless, the poll says what it says, so that feeds into my projections. Brexit also make a very small re-emergence here after being at 0% with everyone since January.
- SNP – 57% (+3 / +10)
- Conservative – 20% (-3 / -2)
- Labour – 14% (+2 / -9)
- Liberal Democrat – 6% (-2 / -2)
- Green – 1% (-1 / nc)
In part because it is undeniably dramatic and in part because we’re still really bad at dealing with Proportional Representation in Scotland, this is the side of the poll which has seen most coverage. Whilst telling a story that’s pretty similar to the List vote poll, for the SNP to be on 57% is astonishing. That’s the highest they’ve been with any pollster since 2016, and exceeded only by an outlying 60% found by TNS a few months ahead of the 2016 election. Everyone else is therefore rather in the doldrums, though again, Labour are at least up compared to the previous poll.
That massive 10% gap between the SNP’s figure in each poll is quite different to Panelbase last month, where the difference was just 5%. We haven’t had a Survation for a while, but they also tend to find wider gaps when they do poll. Survation mention the list vote is proportional, whilst neither YouGov nor Panelbase do, so I’m left scratching my head as to what else about the methodology might prompt such differences.
Anyway, if we do the usual projection into seats, it might look like;
- SNP – 74 (+6 / +11)
- Conservative – 29 (+1 / -2)
- Labour – 17 (+2 / -7)
- Liberal Democrat – 5 (-2 / nc)
- Green – 3 (-7 / -3)
- SSP – 1 (nc / +1)
That’s another huge SNP majority, whilst everyone currently in parliament except the Lib Dems drop seats, with particularly steep drops for Labour and the Greens. Although the Lib Dems and Greens have the same vote share, remember the regional rather than national nature of proportionality under AMS. When both parties are at 6% nationally as they are here, the Lib Dems outperform the Greens per my calculator. When both are at 7%, it’s the Greens that come ahead. In short, AMS is much more sensitive to both small national shifts and quirks of vote distribution than stronger forms of PR. Note too that as with the last poll, that remarkable SSP figure translates to a Glasgow seat.
Westminster Voting Intention (Tracker)
- SNP – 54% (+3 / +9)
- Con – 20% (-5 / -5)
- Lab – 16% (+1 / -3)
- LD – 5% (-1 / -5)
- Grn – 2% (nc / +1)
- Brex – 2% (+2 / +1)
As you’d expect, Westminster voting intention looks very similar. This is only the sixth Scotland-only Westminster poll since the December election, but this is again the highest figure for the SNP either YouGov or Panelbase have found in that time. On the other side, the Conservatives and Lib Dems are at their lowest, though in the latter’s case they’ve had this figure in half of those polls. Brexit also make a small re-emergence here to come out evens with the Greens.
Constitutional (Independence Tracker)
Changes for this one are versus the 22nd – 27th of January, as the April poll didn’t contain any Independence figures.
- Yes – 45% (+2)
- No – 40% (-2)
- Don’t Know – 9% (-1)
Yes further increasing its lead over No here, though it’s still a minority view. Note that YouGov have larger refused/wouldn’t vote figures than other pollsters, which I don’t report here for consistency, which deflate the Yes and No figures slightly. If we exclude Don’t Knows and take it down to pure Yes-No;
- Yes – 53% (+4 / +8)
- No – 47% (-4 / -8)
This continues the trend of Yes leads in polling that we’ve seen all year. For the moment, it seems the new normal is that more Scots support Independence than do the Union, though it’s not a very large lead and almost certainly being driven in part by the ongoing pandemic. If nothing else however, this question has become much more interesting than it has been for years.
In addition to the actual Yes/No question, there were questions about attitudes towards holding another referendum next year. In the event of an SNP majority, should there be another referendum:
- Should – 44% (+3)
- Shouldn’t – 41% (-5)
- Don’t Know – 15% (+2)
So, in the event the SNP have a majority, a slender plurality of Scots think that should lead to another referendum next year.
If the SNP don’t have a majority but there is an overall majority in parliament (wasn’t asked in Jan):
- Should – 39%
- Shouldn’t – 46%
- Don’t Know – 15%
Interestingly, there’s a stronger sense there shouldn’t be one if there’s a Pro-Independence majority that isn’t solely SNP at Holyrood. That may suggest that despite the long presence of Greens and the, notional, SSP support in this poll, Independence remains overwhelmingly associated with the SNP.
And if there’s a majority against:
- Should – 25%
- Shouldn’t – 58%
- Don’t Know – 17%
Unsurprisingly, a large majority of Scots don’t want a referendum if a majority of their parliamentarians don’t want one either.
Moving beyond next year, there was also a flat in the next 5 years question:
- Should – 47% (+3)
- Shouldn’t – 37% (-2)
- Don’t Know – 16% (-1)
Although not quite a majority, there’s a reasonably strong sense amongst voters that the next 5 years would be an appropriate time for a referendum.
Finally on this, what about the eventuality that there is an SNP majority at Holyrood, but the UK Government won’t transfer the powers to hold a referendum? Should the Scottish Government try and hold one anyway:
- Should – 38%
- Shouldn’t – 44%
- Don’t Know – 18%
No majority but a clear lead for people opposed to the Catalonian idea of holding a referendum without the UK Goverment’s assent.
There were also a batch of approval ratings in this poll. These are ratings of performance as Very/Fairly Well/Badly, so for character saving I’m going to display these so the figures in brackets are Very/Fairly.
Starting with the First Minister, Nicola Sturgeon:
- Well – 72% (38% / 34%)
- Badly – 22% (12% / 10%)
- Don’t Know – 6%
Then Prime Minister, Boris Johnson:
- Well – 20% (2% / 18%)
- Badly – 74% (44% / 30%)
- Don’t Know – 6%
Completely opposing figures for the two key leaders, with Sturgeon holding a net approval of 50% versus -54% for Johnson.
New Scottish Conservative Leader, Douglas Ross (this is will he, not is he):
- Well – 17% (4% / 13%)
- Badly – 32% (20% / 12%)
- Don’t Know – 51%
Unsurprisingly, given he’s only just entered the scene, most people either don’t know who Douglas Ross is or what to make of him. Those who do already give him a net negative rating of -15%.
At Holyrood, he’s being subbed for by former Leader Ruth Davidson:
- Well – 43% (17% / 26%)
- Badly – 28% (14% / 14%)
- Don’t Know – 29%
By contrast, Davidson remains much more well known and still broadly popular, with a net positive rating the reverse of Ross at 15%.
Scottish Labour Leader, Richard Leonard:
- Well – 10% (1% / 9%)
- Badly – 37% (18% / 19%)
- Don’t Know – 53%
Despite having been in post for approaching 3 years, Leonard is even more of a mystery than Ross is, and maintains a net rating of -27% on top of that.
UK Labour Leader, Keir Starmer:
- Well – 39% (6% / 33%)
- Badly – 25% (10% / 15%)
- Don’t Know – 36%
Leonard’s UK counterpart is more well known and also fares comparatively well amongst those who have an opinion, with a net positive of 14%.
Finally, we’re onto those bits and pieces I like to do to illustrate more proportional election scenarios. Starting with the BBS-favoured “Scandinavian Style” electoral system (changes vs AMS projection / vs last poll):
- SNP – 62 (-12 / +4)
- Conservative – 28 (-1 / -2)
- Labour – 18 (+1 / +2)
- Liberal Democrat – 8 (+3 / -1)
- Green – 8 (+5 / -3)
- SSP – 5 (+4 / nc)
Without a majority of the vote, in a maximally-proportional system the SNP end up just shy of a majority of seats. Substantially boosted Green and SSP groups would still keep the balance in favour of Independence, however.
(Note that the “last poll” comparison doesn’t match the figures in April’s post. I’ve since changed the allocation of direct seats from D’Hondt to Sainte-Laguë because the SNP were doing so well it was disrupting proportionality. This reflects change under the new model.)
Giving Westminster the same treatment (changes only vs last poll, as I don’t do pure FPTP projections):
- SNP – 34 (+2)
- Conservative – 12 (-3)
- Labour – 10 (+1)
- Liberal Democrat – 3 (-1)
Since the SNP do have a majority of the Westminster vote, they’d come out with a majority of seats. Nonetheless, this would be a vast improvement on the reality of FPTP, which on these figures folk typically expect to tip everything except Edinburgh Southern and possibly Orkney and Shetland the SNP’s way. As ever, the mind boggles that anyone would think FPTP is a democratically defensible system.
The final piece returns to Holyrood, with a range of alternative systems that further illustrate how lacking AMS is when it comes to accurately reflecting the views of voters.