Poll Analysis: Savanta 15th – 17th of February 2023

For the reasons outlined in the introduction to this piece, Ballot Box Scotland was supposed to be on a break from Twitter, focussing primarily on the website and even then running shorter form analysis than usual of polls. Then the First Minister announced she was standing down. As much as I may wish otherwise, it’s not really possible just to disengage entirely when probably the biggest political event in Scotland since the referendum has just taken place. So I’m back-ish, earlier than I’d have liked.

With three polls having dropped on Friday (the 17th of February) and a fourth the next day, that would have been a polling overload at the best of times, never mind when I’m a bit knackered, have a day job, and had social plans for most of the evening. I know I always complain about the relative dearth of Scottish polling but c’mon, a little breathing room would help! Given both my previously stated intention to keep these pieces short and how many polls there are to get through anyway, I’m going to rush pretty quickly through them.

This third poll comes from the usual duo of Savanta (link to tables) and the Scotsman (link to original writeup). It’s also the first poll to have been conducted since Sturgeon’s announcement, going into the field on the very day, and so gives us our first glimpse at how things might begin to shift (if at all) in this new era. At least, it would have, if there wasn’t another poll out covering the same period with some notable differences. If opinion polling is good for one thing, it’s leaving us all completely bemused – and you can take that as a professional opinion in my case.

The previous Savanta covered the 16th – 21st of December 2022. Changes are shown as (vs that poll / vs last election).

Regional Vote

Well, initially, it doesn’t seem like that much effect – indeed, most of the movement seems to be from Conservative and Lib Dem to Labour, leaving the SNP with another quite slender lead. The Conservative figure is the one that interests me most here, as Savanta have tended to have them stubbornly higher than other pollsters even through their worst weeks last year. They appear to have fallen in line on that front now. Otherwise, the Greens and Lib Dems have done a little trade, with Greens back up to a joint-record share, whilst Lib Dems slip very marginally from a joint-record this term share to a still very positive figure.

Constituency Vote

Broadly similar(ly quiet) over on the constituency ballot, though with margin of error swings going in Labour’s favour. That actually gives them their best constituency poll of the term, and indeed since the very beginning of 2015.

Seat Projection

Projecting that into seats might give us something like this:

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

Although the SNP still had a significant lead over Labour on the constituency vote, Labour hitting their best figure yet helps narrow things enough to tip a handful of extra constituencies. It’s not quite as many as in the pre-resignation YouGov, as the SNP vote is higher in this one, but it’s still enough to bring the SNP down to the same term thus far low of 55 seats.

Counterbalancing that, in part because losing constituencies lessens the overhang pressure, with a doubling of their ranks this is the strongest projection for the Greens in their history. That effectively leaves the joint SNP-Green government where is is now, with a comfortable majority. Note that the Greens again have Kelvin here, despite not showing up in the constituency side of the ballot – my model drives Green figures partly via the list, just because their polled constituency vote is useless for those purposes, and they obviously won’t get 0% in Kelvin!

The Westminster figures might look quite boring as well, especially compared to the YouGov, but this is what I would refer to as the beginning of the danger zone for the SNP: being within 10% of Labour. As noted in the YouGov piece, the SNP have a well spread vote, and Labour have a more concentrated vote. That means Labour are likely to be significantly above 32% in the urban central belt, and therefore would rapidly begin to flip seats from the SNP. Predicting FPTP is a tricky business (I only do it as part of Holyrood because of the moderating impact of list seats) but this is where I would start to say Labour could enter double-digit seats, even if only just.

Absolutely zero change versus the last poll here, another very close to tied result with a narrow edge for the Union. Given this is post-Sturgeon’s resignation, there’s a little warning here: don’t assume the departure of Sturgeon is the death of Independence. I keep making the point that we’re stuck on the constitution because we aren’t, really, discussing the substance and instead engaged in a bunch of procedural proxy debates.

That point had a helpful little addition, I thought, from Professor Aileen McHarg here, remarking that both sides have relied a bit on “luck”, getting events outside of their own control that seem to boost their side. But those events aren’t a substitute for actually properly addressing the constitutional question and putting a case to people. Shouting “No to IndyRef 2!” is not a governing prospectus, never mind a reason to back the Union, for example! Similarly, growing support for Independence needs more than saying it’s undemocratic for the UK Goverment not to allow a referendum.


As ever, the last little bit of analysis concerns those hypothetical and more proportional voting systems that BBS likes to play about with. The use of pure FPTP at Westminster 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.

Although this poll has the SNP on the same number of seats under the ordinary projection as the YouGov, the significantly better vote share for the Greens re-asserts the general tendency of this more proportional alternative to show a weakened but still majority SNP-Green government.

That can’t withstand the almost maximally proportional system however, which gives a lead for the combined Pro-Union parties.

Scandinavian Style Westminster

Although the SNP are at the precipice of what I’ve dubbed the danger zone under FPTP, it’s still nowhere near close enough to deprive them of a majority of MPs. Bring in a proportional system, and things shift significantly – mostly in the Conservative’s favour, given how badly they’d do under FPTP, but also in Labour’s.

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)