Poll Analysis: Survation 20th – 22nd of April 2021

Keep tabs on all the latest polling, articles and information ahead of the 2021 Scottish Parliament election in the Ballot Box Scotland Holyrood Hub!

After a very busy month for Survation in March, where they released a whopping three polls, they chilled out a bit and let us go most of April without another. The fieldwork for this one takes us up until two weeks before the election, and polls will be picking up increasing numbers of people who have actually already voted by post. We’re in the home stretch now, and parties have very little opportunity left to deliver big turnarounds.

Display format for this post:

Regional Vote

SNP ~ 35% (-3 / -7)
Labour ~ 22% (+3 / +3)
Conservative ~ 20% (+2 / -3)
Green ~ 10% (-1 / +3)
Liberal Democrat ~ 7% (-1 / +2)
Alba ~ 3% (nc / +3)
Reform UK ~ 1% (nc / +1)
UKIP ~ 1% (nc / -1)

Though the SNP maintain their clear lead, a sizeable dip down to 35% gives them their worst figure for this vote since March 2019. That leads to a whopping 15% gap between their two ballots, which I can’t help but feel is likely to close from one end or the other on the day itself. Labour and the Conservatives both make gains on the last poll, with Labour on their best share since December 2018. That comes, as ever, with the massive warning that Survation are the only pollster finding Labour doing this well and consistently placing second ahead of the Conservatives.

For the smaller parties, there are slight decreases for both the Greens and Lib Dems, though that keeps them on the upper end of their recent polling figures, and Alba remain steady on 3%. As I’ve noted previously, that’s roughly the most uncertain point. It’s conceivable they could squeak Alex Salmond in with that on the assumption their vote is strongest in the North East, but it would need to be heavily concentrated.

Constituency Vote

SNP ~ 50% (+1 / +3)
Labour ~ 21% (+1 / -2)
Conservative ~ 21% (+1 / -1)
Liberal Democrat ~ 7% (-2 / -1)
Green ~ 1% (nc / nc)

In contrast to their recent low for the list vote, the SNP are actually up very slightly in the constituency. So too are Labour and the Conservatives, all of this coming at the Lib Dems’ expense.

Seat Projection

Please see this page for how projections work and important caveats.

Projecting that into seats might give us something like this:

SNP ~ 67 (+1 / +4)
Labour ~ 25 (+2 / +1)
Conservative ~ 21 (nc / -10)
Green ~ 10 (-1 / +4)
Liberal Democrat ~ 6 (-2 / +1)

A big constituency lead for the SNP projects to a majority on constituencies alone, with nary a list seat to be had. This is really stretching the intended functionality of AMS by this point, with a Gallagher Index of 13 representing a highly disproportionate result. Overhang really is absurd when you have that degree of difference between the two votes!

SNP ~ 47% (-2 / +2)
Labour ~ 22% (+1 / +3)
Conservative ~ 21% (nc / -4)
Liberal Democrat ~ 8% (nc / -2)
Green ~ 1% (nc / nc)

No Westminster Q in this one.

Standard Question

No ~ 45% (+1)
Yes ~ 44% (-1)
Don't Know ~ 11% (nc)

At this point, regular readers should be able to reel off, word-for-word, what my analysis of this is. It’s on a knife-edge, totally margin of error, who knows which way Scotland would go in an actual referendum right now? The last poll had Yes narrowly ahead here, this time it’s No narrowly ahead.

Excluding Don't Knows

No ~ 51% (+1 / -4)
Yes ~ 49% (-1 / +4)

Though the last poll was a tie with the reverse Yes-No figures, this time the balance of responses tips it just towards a 51-49 split in favour of the Union. Again, however, we really have to emphasise just how much this lies within the margin of error. If you think that’s unsatisfying, try being the person who has had to write that two or three times a week every week for the past month!

Council Area Projection

Please see this page for how projections work and important caveats.

On a simple Uniform Swing versus 2014, this might put a total of 20 Councils in the No column, a decrease from 28 in 2014. Yes could be on 12, up from 4 at the referendum.


As ever, the last little bit of analysis concerns those hypothetical and more proportional voting systems that I have a bee in my bonnet about here at BBS. The fact Westminster uses pure FPTP is an affront to democracy, and though Holyrood fares far better, AMS is still deeply imperfect. The examples here simply transpose the poll findings onto more proportional voting systems – the reality is that different systems would of course result in different voter behaviour.

As the election approaches and polls are coming out more frequently, I’m jettisoning the actual map for this one for now. I still think it’s a fun and useful exercise, but it’s also time consuming to manually change 129 bubbles! (The AMS projection is only 56 bubbles, because the 73 constituencies are driven from a spreadsheet).

Changes here are vs AMS / vs same projection for the last poll.

SNP ~ 46 (-21 / -3)
Labour ~ 29 (+4 / +3)
Conservative ~ 27 (+6 / +3)
Green ~ 13 (+3 / -2)
Liberal Democrat ~ 10 (+4 / -1)
Alba ~ 4 (+4 / nc)

This really emphasises just how disproportionate AMS is on this poll projection, as the SNP would be down a staggering 21 seats if they are apportioned entirely proportionally to the list vote. Even with the Greens and Alba, the pro-Independence camp would only total 63 seats, versus 66 for the combined pro-Union parties.

Scandinavian Style Westminster

Changes here are only versus the same projection on the last poll, given my aversion to doing projections on pure FPTP.

SNP - 28 (-1)
Labour - 13 (+1)
Conservative - 13 (nc)
Liberal Democrat - 5 (nc)

Shorn entirely of the distorting effect of FPTP, Scotland’s Westminster delegation would look substantially more like the country’s electorate. The SNP would still be pretty dominant, just not to the artificial level they currently are.

If you find this or other Ballot Box Scotland output useful and/or interesting, and you can afford to do so, please consider donating to support my work. I love doing this, but it’s a one-man project and takes a lot of time and effort. All donations, no matter how small, are greatly appreciated and extremely helpful.
(About Donations)