Survation, 28th Sept-2nd Oct 2018

Once again Scottish polls prove just like buses – we’d heard nothing since July, then three came along at once. That probably had something to do with it being SNP Conference at the weekend. Indeed, one of the three was commissioned by the SNP themselves, but more on that later. I’ll cover these in order, so starting with the Survation poll from the 28th of September to the 2nd of October.

Display format;

Holyrood Voting Intention and Projection (Tracker)


  • SNP – 43% (nc / -4)
  • Conservative – 24% (nc / +2)
  • Labour – 23% (+2 / nc)
  • Lib Dem – 9% (nc / +1)
  • Green – Not prompted for


  • SNP – 32% (-1 / -10)
  • Labour – 23% (+2 / +4)
  • Conservative – 21% (+2 / -2)
  • Green – 10% (-1 / +3)
  • Lib Dem – 8% (-2 / +3)
  • UKIP – 5% (nc / +3)

Not a whole lot different compared to July’s poll, a small bump for Labour and the Conservatives with similarly small drops for the SNP, Lib Dems and Greens. UKIP also hold steady on that remarkably high 5% Survation found last time.

Basically, this poll keeps with Survation’s general pattern of late. It’s still the only pollster showing Labour ahead of the Conservatives on the list vote, the only one showing high UKIP figures, and consistently the best Green and Lib Dem results.

If we project that into seats it might look like;

  • SNP – 59 (nc / -4)
  • Labour – 27 (+1 / +3)
  • Conservative – 25 (+1 / -6)
  • Green – 10 (nc / +4)
  • Lib Dem – 8 (-1 / +3)
  • UKIP – 0 (-1 / nc)

Two notable aspects to this projection. Firstly, Survation again buck the general trend amongst pollsters by suggesting a pro-Independence majority. Secondly, although the previous 5% poll had UKIP with a seat in the South, they come up empty handed here.

If UKIP did genuinely win 5%, I expect they would win a seat or two, it’s just that modelling based on 2016 results doesn’t tend to think so, thanks to that overhang phenomenon I keep mentioning. Effectively the improved vote share for Labour lifted them enough to deny UKIP the last South seat in this projection.

It’s also worth pointing out how vital that list vote is to Labour overtaking the Conservatives, who polled marginally ahead in the constituency vote. On these figures the SNP are benefiting massively from a constituency overhang, which ends up counting in Labour’s favour since they aren’t really in the running in many constituencies the way the Tories are.

As with the previous Survation poll, the only constituency seat projected to flip is Perthshire South and Kinross-shire – but still, plenty of close calls.

Westminster Voting Intention (Tracker)

SNP – 41% (-1 / +4)
Conservative – 27% (+3 / -2)
Labour – 24% (+1 / -3)
Lib Dem – 6% (-2 / -1)
Green – Not prompted for

A bit of a dip for both the SNP and Lib Dems whilst the Conservatives and Labour are up, but still suggesting the SNP would come out better than they did in last year’s snap election.

Independence Voting Intention (Tracker)

  • Yes – 43% (+1)
  • No – 49% (+1)
  • Don’t Know – 8% (-2)

Once Don’t Knows are excluded;

  • Yes – 47% (nc / +2)
  • No – 53% (nc / -2)

No particular surprises, based on recent polling anyway, to find the split on independence is still almost exactly where it was in 2014. They also asked a follow up about the impact of Brexit – “Does the prospect of the UK leaving the EU make you more or less likely to support Scottish Independence?”;

  • No more or less likely – 36%
  • More likely – 35%
  • Less likely – 20%

So although Brexit isn’t leading to the surge in support for independence many supporters may have wished for, it is causing more people to lean towards independence than it is putting off.

Brexit Voting Intention

The poll also included a few Brexit related questions that didn’t tie in with independence. On a straight-up re-run of the 2016 referendum;

  • Remain – 62%
  • Leave – 33%
  • Don’t Know – 5%

Once Don’t Knows are excluded;

  • Remain – 66% (+4)
  • Leave – 34% (-4)

Confidence in UK Government securing the best possible Brexit deal;

  • A lot – 5%
  • Some – 20%
  • Not much – 30%
  • None – 40%
  • Don’t know – 6%

If current “impasse” in Brexit negotiations can’t be resolved by March 2019;

  • Leave with no agreement – 29%
  • Shouldn’t leave with no agreement – 51%
  • Don’t know – 19%

When the negotiations are complete, support or oppose a referendum on the deal;

  • Support – 49%
  • Oppose – 29%
  • Don’t know – 22%

If there is such a referendum, should there be a remain option;

  • Should – 55%
  • Should not – 32%
  • Don’t know – 13%

Summing all of those various questions up, Scotland is still overwhelmingly in favour of EU membership, has little confidence in the UK government’s ability to negotiate a good Brexit deal, doesn’t want to leave without one, on which it wants to see a referendum with the option to Remain.

That’s quite a shopping list of items the UK Government has so far set themselves dead against.

That SNP Survation Poll…

There were two Survation polls released at the weekend there. As mentioned at the start, one of them was commissioned by the SNP. For the time being, I won’t be including that poll anywhere on this website. There was actually some interesting stuff in it, but I’m viewing the Holyrood List vote portion as being so dubious I’m going all-out on letting the poll pass rather than cherry pick what I do and don’t show from it.

Basically, their list vote findings were so out of kilter with Survation’s own trends I have no idea how they came to them, even though they sit largely within the margin of error. The SNP come out at 40%, far in excess of the 33% that the same pollster found two days before, and which is in line with all other agencies. By contrast, the Greens 4% is unprecedented since 2016, and flies in the face of the 10%. UKIP don’t even register individually, with “Other” only coming in at 1%, despite the already discussed tendency of Survation to show favourable UKIP figures. And it’s the only Survation list poll to put the Conservatives ahead of Labour.

If that had come as an isolated poll, my eyebrows would have been raised, but I’d have duly put it all neatly in the appropriate places. Coming days after the poll reported here, it jars so badly that unless a further poll (from Survation themselves or any other agency) comes close to showing similar figures, it’s effectively sitting in quarantine.