Poll Analysis: Panelbase 2nd – 5th of October 2023

In the first poll published since the Rutherglen and Hamilton West by-election, but concluding before it, we’ve got the common pairing of Panelbase (link to tables) and the Sunday Times (link to original writeup). This one squeezed itself out just in time for it to be after I’d said I was taking a week off Ballot Box Scotland, but before my actual time away from home began. I’ve therefore pushed the coverage out for it as quickly as possible (hence not waiting for tables link), because I’d rather not having it waiting on my to-do list after a little holiday. Note that I will be away Monday the 9th through most of Friday the 13th (spooky!) and thus if any more polls come out in that time I will not be able to cover them.

This is hopefully the last time I will make the reference, but folk may recall I made a big song and dance about not giving Panelbase/Sunday Times polls the full suite of BBS coverage as they’d fallen into the habit of not reporting the whole poll at the same time. Indeed, for a few releases, the Holyrood figures hadn’t even made it into the published tables. I don’t know whether it was my kicking up a stink that did it, but this is now third in a row where they’ve released everything all at once. I’m genuinely pleased that’s the case, as I don’t like skipping polls, not least because people get hopping mad and accuse me of bias when I do.

The previous Panelbase covered the 12th – 15th of June 2023. Changes are shown as (vs that poll / vs last election).

Regional Vote

Everything on the list vote is within margin of error, but within that Labour are up a couple of points to hit a joint-best on this vote (in polls that BBS tracks) since the referendum. That puts them a point ahead of the SNP, in a thus-far very rare lead. The Conservatives also tick up a little bit, whilst the Panelbase tendency to be more favourable to Alba than any other pollster seems to have eroded somewhat. Neither the Greens nor Lib Dems are changed, which means Panelbase are really cementing themselves as a highball pollster for the Greens after a long history of lowballing them.

Constituency Vote

Unlike the list vote, the SNP remain ahead at constituency level, but again have very slight further reduction, which puts this as their worst share of the term in any poll I’ve reported for BBS. Labour are nipping at their heels but unchanged, with the big shift in this vote being in the Conservatives’ favour, which may be more of a corrective for what was a term-low last poll. Dropping 2% puts the Lib Dems on par with the Greens, who remain at their best ever polled constituency share.

Same point I made about the Greens last time stands; it’s hard to get this figure “right” as Green constituency performances where they do stand suggest standing everywhere would lead to significantly higher share than the lowballs other pollsters give, but the fact they don’t stand everywhere makes the Panelbase figure look suspect.

Seat Projection

Projecting that into seats might give us something like this:

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

Relative to the last poll, the SNP further continue their slide whilst their partners in government remain static. A large part of what catches the SNP here is actually the Conservative recovery on the constituency side of things. Although they only have a net gain of one, the SNP were overhanging enormously in the North East last poll, and the Conservatives picking up two of those constituencies here eases that pressure. As the Lab-LD bloc is still smaller than SNP-Green, Labour would likely actively need Conservative support rather than simply abstention to take office in the first place, but to deliver their agenda this scenario would offer a bare-majority 65 seats to the Lab-LD-Green traffic light option.

I don’t normally point out differences between my projections and others’, but just because in this case it’s the difference between “Labour lead in seats” (per the projection given by Professor John Curtice to the Sunday Times) and “Labour and SNP tie in seats” (per my calculator) I’ll quickly do so. Effectively, my calculator gives two seats more for the Lib Dems than the good Professor’s, which come one apiece from Labour and the Greens. Such small changes are to be expected as projecting isn’t an exact science, and aren’t a sign of anything other than small methodological differences.

Perhaps oddly, although once again I’m having to make the caution about margins of error, the SNP actually gain on this vote. That gives them a tiny bit of breathing space ahead of Labour, who they’d been tied with in the last poll, whilst the Conservatives are unchanged and the Lib Dems up a point. I continue to find it odd that Panelbase aren’t prompting for Greens on this question, since they are finding such high shares for the Holyrood constituency vote. It seems likely they’d be on at least the 3% I use as my proportional model threshold if they were prompted.

Despite the SNP lead here, as I’ve noted a few times this is the kind of national margin between them and Labour that could, when you consider Labour would be ahead in the Central Belt on these figures, see Labour win more seats overall. FPTP is a voting system where you benefit from a concentrated rather than a spread vote share. On these kinds of figures that could present a real difficulty for the SNP and an opportunity for Labour.

Margin of error shifts on the constitutional question with the overall result likewise within margin of error, leaving things stubbornly unchanged? Groundbreaking. Devil Wears Prada adaptations aside, polling continues to show what it has for a little while now: though net support for the Pro-Independence camp has reduced this year, that’s not really the case for Independence itself. It remains in the minority but partisan switching isn’t leading to constitutional switching – at least, not yet.


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.

Scandinavian Style Westminster

Note that this is the first Panelbase since the final Westminster boundary changes were submitted, so the figures here sum up to 57 rather than 59. I’ve therefore recalculated both the previous poll and the 2019 result to match 57 seats. Remember that Panelbase are rather oddly not prompting for the Greens at Westminster there’s no Green presence here, though per my note earlier I’d imagine they’d be on at least 3% and thus on a couple of seats here. (At bang on 3%, it’d be one from each of the SNP and Conservatives).

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)