Poll Analysis: Panelbase 17th – 19th of August 2022

I’m afraid that at Ballot Box Scotland, I’ve run out of patience for polling that the Sunday Times is commissioning from Panelbase. I want to be clear at the outset the following criticisms lie entirely with the Sunday Times, and not with Panelbase. When it comes to commissioned polling, members of the British Polling Council must publish tables of the data within two days of it entering the public domain. However, whether it enters the public domain is entirely down to whoever commissioned it, and they can choose to do so piecemeal. 

This year, the Sunday Times have clearly taken an editorial decision not to put Holyrood voting intention data from their polls into the public domain. At the time of writing, the Holyrood figures for this poll do not appear to have been reported by the Sunday Times (or its weekday counterpart the Times), and thus they are not included in the tables from Panelbase. I am only aware of the headline figures as they have appeared on What Scotland Thinks, a website associated with the UK’s pre-eminent polling expert (and every polling nerd’s hero), Professor John Curtice.

Holyrood data from the previous poll was only added to the tables upon my querying with the paper and with Panelbase two weeks ago – only Panelbase replied. You’ll see if you follow the link to the analysis piece that although it’s backdated, it was only published on the 28th of August and is thus a “lite” entry. Following the link to the previous poll again will take you to another “lite” Panelbase poll analysis, two weeks after the initial findings of the poll had been published.

Once again following the link back will take you to a poll in November – but wait! It turns out that’s NOT the previous poll. Again, data only added to Panelbase tables on the 22nd of August reveals that there was Holyrood data in a poll from the 4th of February. I had to re-do a fair bit of the data on my tracker page as a result of the emergence of this poll months after the fact.

In summary, the Sunday Times have commissioned four polls from Panelbase this year, from which they published some data in their paper – typically Independence and Westminster voting intention, plus topical issues. On none of these occasions have they reported the Holyrood figures from the same poll at the same time, and in most cases the data was not cleared for publication in table form on Panelbase’s website either.

This is, frankly, pathetic. We are 23 years into devolution and I cannot fathom why anyone would spend thousands of pounds on Scottish political polling only to ignore the findings relating to the Scottish Parliament. Whilst I recognise the purpose of commissioning polling is to drive traffic and subscriptions to their paper, not to make life easy for polling aggregators using the data for free, this is an indefensible approach from one of the UK’s major papers. I’ve therefore decided to apply what little weight Ballot Box Scotland has to the situation.

Subsequent to this poll, Ballot Box Scotland will no longer be publishing Twitter updates or website analysis pieces for polls conducted by Panelbase for the Sunday Times.

This policy will remain in place unless and until the Sunday Times begin reporting Holyrood findings in full and at the same time as Westminster and Independence figures. Figures will continue to be added to the relevant tracker pages, as it would be a distortion of the political reality in Scotland to ignore them entirely.

The previous Panelbase covered the 29th of June – 1st of July. Changes are shown as (vs that poll / vs last election).

Regional Vote

This poll is all about the ties. Labour appear to have dipped a bit since the last poll, putting them on a par with the Conservatives who’ve had a very small increase. Similarly, the Greens and Lib Dems are tied, with the former recovering from their worst poll of the term so far. Regardless of what’s happening below them, the SNP remain clearly out in front, though their swing of -3% is the largest of any party and the only one pushing at the edges of margin of error.

Constituency Vote

There’s more separation between the parties over on the constituency with Labour maintaining a clear second over the Conservatives in third, with the latter again seeing a bit of a bump. Both the SNP and Lib Dems have slight decreases versus the last poll. I expect 3-4% of the “others” to be for the Greens, given they had 3% in the previous poll, but we won’t know unless the data is added to the tables.

Seat Projection

Projecting that into seats might give us something like this:

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

Despite the tie in votes, the Conservatives come off slightly worse in terms of where those votes are and thus place a seat behind Labour. However it’s totally even between the Greens and Lib Dems, giving the latter their best seat projection of the term thus far. It’s notable that’s despite it not being their highest regional vote share – effectively, the comparatively weak position of the other two pro-Union parties relative to recent polls creates enough spare for them to squeak through.

Overall, the two Green gains would offset two SNP losses, and the co-operative government between the two would be re-elected.

A lot of the same dynamics – or lack thereof – over on the Westminster vote. The SNP are down a bit, but so too are Labour and the Lib Dems, whilst an increase for the Conservatives still leaves them down a fifth on their 2019 result. There’s little to read into the Green figure here, even beyond the fact there rarely is for Westminster, as they weren’t disaggregated from “others” in the previous poll.

Some slight changes here relative to the last poll, which puts the Union back in the lead on the headline figures, after a wafer-thin 1% lead for Independence. That means what was a tie after removing don’t knows breaks in the Union’s favour this time. Nonetheless, this is still very much on a knife-edge, and you can imagine my usual spiel about being able to predict very little from this.

Timing of a Referendum

Rather than ask about specific timescales, including whether people supported a vote next October, Panelbase appear to have simply asked whether there should be a referendum within the next five years. As with Independence itself, Scots are pretty evenly divided on this, though those who would like to see another vote have a slight lead.


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.

For the moment, although the maps are useful for illustrative purposes, I’m opting just to show these hypotheticals as charts. It’s very time consuming making maps, and for these pure hypotheticals, it’s possibly a bit overkill.

Although it reduces the relative strength of the SNP, a reformed version of AMS actually ends up quite similar to what we currently have. That’s a tie between the Greens and Lib Dems, a single-seat advantage for Labour, and a clear SNP-Green majority…

But turning to a more fully proportional system results in both pairs of tied votes leading to tied seats, whilst returning a relatively slender 66:63 majority for the pro-Union parties.

Scandinavian Style Westminster

Bringing solid proportionality to Westminster naturally breaks the SNP’s dominance, allowing accurate representation for Labour and the Conservatives. The Greens even make it through in this model, having just made the threshold.

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)