Panelbase, 28th-31st Jan 2020

The second of Monday’s double polling whammy was from Panelbase. They’d polled a lot more recently than Survation so in that sense it’s less exciting to see them, but in other ways this one is more dramatic. If you don’t remember seeing my post about the November poll, don’t worry – it took Panelbase so long to release the tables for it that, being totally absorbed in the snap General Election, I forgot to actually post the draft when tables were finally released! It is now available, however, backdated to the appropriate point.

Display format;

Holyrood Voting Intention and Projection (Tracker)


  • SNP – 47% (+8 / +6)
  • Conservative – 25% (nc / +2)
  • Labour – 14% (-3 / -5)
  • Green – 7% (+1 / nc)
  • Liberal Democrat – 7% (-2 / +2)
  • Brexit – 0% (-3 / nc)

Massive gains for the SNP here on both the previous poll and 2016, which given this is the proportional element of the system really matters. Labour are the big losers, still suffering after their historic low at the snap election. The Lib Dems too, whilst up on their 2016 result, are continuing to slide back downwards as their EU election revival seemingly recedes. Basically, compared to the Survation poll, Panelbase comes out less rosy for the three smaller parties at Holyrood, and better for the SNP and Conservatives. Note too that Brexit appear to have totally disappeared from view – there is a 2% for “others” listed (yes, that then sums this up to 102%, effect of rounding) which at a guess I’d say would mostly be Brexit.


  • SNP – 50% (+7 / +3)
  • Conservative – 26% (nc / +4)
  • Labour – 14% (-3 / -9)
  • Liberal Democrat – 7% (-2 / -1)
  • Green – 3% (+1 / +2)
  • Brexit – 0% (-3 / nc)

It’s mostly the same story on the Constituency vote, but unlike Survation, there isn’t a massive gap between the SNP votes on each ballot. The Conservatives aren’t any different to the last poll, but they would gain a fair bit versus 2016. Labour and the Lib Dems are both down versus 2016 as well as the previous poll, though it’s only a marginal dip for the latter.

Someone on Twitter has suggested the stark difference in list vote findings between the two polls may come down to the phrasing of the question. Survation’s question uses “second vote” terminology which, the suggestion goes, could imply it’s a second preference vote. That’s a sensible idea, but I’m not sure it’s entirely right. The Survation question actually specifies that the regional vote is proportional, whereas Panelbase doesn’t. I agree that it’s likely that phrasing accounts for at least some of the difference (not all, given the Constituency figures are also rather different), but probably because it’s emphasising the vote is proportional rather than implying it’s a second preference.

Anyway, methodology musings aside, translating this into seats might look like;

  • SNP – 68 (+8 / +5)
  • Conservative – 32 (-2 / +1)
  • Labour – 17 (-5 / -7)
  • Green – 7 (+3 / +1)
  • Liberal Democrat – 5 (-4 / nc)

Compared to the Survation poll, the SNP’s much closer alignment on the two votes projects here into a majority. Bearing in mind they won 69 seats with 45%/44% in 2011, it shouldn’t be surprising that 50%/47% is pretty certain to deliver one. That alignment also means this projection features one of the lowest overhangs we’ve seen in a while, at just one seat (at the Lib Dems’ expense in West), in line with the 2016 (Lab, Mid and Fife) and 2011 elections (Lib Dem, Lothian).

Less overhang there may be, but the exclusively regionalised nature of proportionality under AMS still makes it useful to pop the list vote into an example more proportional system (vs AMS projection);

  • SNP – 62 (-6)
  • Conservative – 31 (-1)
  • Labour – 18 (+1)
  • Green – 9 (+2)
  • Liberal Democrat – 9 (+4)

As this example system is more proportional, no majority for the SNP on 47% – but combined with the Greens, there would still be a pro-Independence majority. The main beneficiary of more proportionality however would be the Lib Dems, whose spread of votes in 2016 isn’t conducive to gaining seats under AMS without more substantial increases in support, at least in these projections – obviously, in an actual election, that spread may change enough to get more MSPs in.

Constitutional (Independence Tracker)

Usual Independence question here;

  • Yes – 49% (+5)
  • No – 46% (-4)
  • Don’t Know – 6% (nc)

This was widely seen as the big news to come out of this poll. Compared to Survation, where it was neck and neck, and YouGov, where a slender Yes lead was built on a large number of don’t knows, Panelbase are close to Yes having a majority in its own right. Therefore, when we exclude don’t knows…;

  • Yes – 52%
  • No – 48%

… We get the cursed EU Referendum figures! As noted in the post about the Survation poll, we’re seeing a bit of a more Independence favourable trend in recent polls. However, it’s far too early to state whether this will continue, or if the pro-Union norm will reassert itself, as happened immediately after that 2016 referendum.

With the SNP government and MPs showing no signs of going anywhere at the moment, one of the main battlegrounds for the Independence issue hasn’t been the “Independence vs Union” question in and of itself, but instead what constitutes a “mandate” for a second referendum. This poll asked a couple of questions on that point. Firstly, it asked whether in light of the SNP’s 48 of 59 seats victory in the snap General Election there was a mandate for a referendum;

  • Yes – 48%
  • No – 42%
  • Don’t Know – 10%

A large chunk of Don’t Knows here, but a substantial plurality of respondents do think that the December election constitutes a “mandate”. A major aspect of the argument around mandate has related to whether Brexit is reason enough for a second referendum, with the key argument against tending to hinge on the terminology of “once in a generation” used by some leading figures in the 2014 Yes campaign. This poll asked what respondents views were on whether it was enough of a reason;

  • Yes – 52%
  • No – 41%
  • Don’t Know – 7%

So an outright majority of Scots do seem to view Brexit as a big enough issue to justify another referendum. Finally, in light of the refusal of the current UK Government to grant a second referendum, a rather dramatic final question asked whether or not the UK was a “fully democratic country”;

  • No longer fully democratic – 49%
  • Still fully democratic – 39%
  • Don’t Know – 13%

And this final question is only just shy of an outright majority. Whether for the Union or Independence, for or against a second referendum in the short term, that’s an unsettling finding. Fewer respondents could bring themselves to affirm their confidence in the UK’s democracy than were in favour of maintaining the Union. There’s a certain amount of dramatic questions get dramatic answers at play here so folk shouldn’t get too het up about this for the moment, but it’s still a potential cause for concern for the Pro-Union side.