2018 in Review – Parliamentary Polling

For the second part of 2018 in Review, let’s take a look at polling for both the UK and Scottish Parliaments. Rather than give a blow-by-blow account of each poll – I’ve already done that – it’s instead worth seeing how we stand now compared to a year ago.

Since it wouldn’t be representative either to take the last poll of the year or to just average out the whole thing, I’ll be comparing the last quarter of this year with the same period last year. This is a bit of a bumper post as Ballot Box Scotland only launched in January. So to compare this year with last I’m having to go through those 2017 figures from scratch rather than just link to a previous post!

Q4 2017

In a break from my normal projection format, we’ll start with Westminster first since it’s a bit less detailed. There were three polls in the final quarter last year, which averaged out at (changes vs snap GE 2017);

  • SNP – 37.3% (+0.4)
  • Labour – 28.3% (+1.2)
  • Conservative – 24.7% (-3.9)
  • Lib Dem – 7.0% (+1.2)
  • Green – 0% (-0.2)
  • UKIP – 0% (-0.2)

A little bit of the shine seemed to have come off the Conservatives immediately after the snap GE, and they were registering a substantial dip in support, with small boosts for the SNP and Labour, who would have pushed their way back into second. Green and UKIP figures for Westminster are always unreliable anyway due to their patchy record of standing candidates, but they weren’t even prompted for in these polls.

Westminster isn’t my focus, so I don’t do projections myself, but using Electoral Calculus (changes vs snap GE 2017);

  • SNP – 31 (-4)
  • Labour – 12 (+5)
  • Conservative – 12 (-1)
  • Lib Dem – 4 (nc)

Although Labour would have had a slight lead in votes, they were projected to come out even with the Conservatives in terms of seats.

Holyrood polls are few and far between until the election begins to loom, so there were only three polls in the last quarter of 2017. Taking the average of those for the constituency (all changes vs 2016);

  • SNP – 40.0% (-6.5)
  • Labour – 26.0% (+3.4)
  • Conservative – 24.3% (+2.3)
  • Lib Dem – 6.7% (-1.1)
  • Green – 0.7% (+0.1)

And for the list;

  • SNP – 33.3% (-8.4)
  • Labour – 24.3% (+5.2)
  • Conservative – 22.0% (-0.9)
  • Lib Dem – 8.0% (+2.8)
  • Green – 8.0% (+1.4)
  • UKIP – 2.3% (+0.3)
  • SSP – 1.0% (+0.5)

Most notably at this point last year, Labour had managed to reclaim their spot as the 2nd party in polling. Although two of the three were from Survation who are the only pollster to consistently put Labour ahead of the Conservatives, the YouGov also had a small Labour lead. Given they lead in Westminster polling too, we can safely say Labour were enjoying a small post-snap bounce.

Translating all that into seats (changes vs 2016);

  • SNP 54 (-9)
  • Labour – 31 (+7)
  • Conservative – 28 (-3)
  • Lib Dem – 8 (+3)
  • Green – 8 (+2)

If you’ve been following polling this year, it won’t necessarily come as a surprise that at the end of last year, the constitutional balance in Holyrood was projected to tip towards the pro-union parties. Labour would have been in line to quite narrowly overtake the Conservatives, and potentially even win the Rutherglen and Cowdenbeath constituencies. Meanwhile the Lib Dems and Greens would have been even with a run of 8’s.

Q4 2018

A year on, how does it look? With Brexit and all kinds of dramas over Continuity Bills, maybe there could have been some major shifts in polling? Aye, nae chance.

For Westminster we have five reliable polls in Q4 2018. Note the change figures here are different from what auto-Tweeted this morning, because guess who had let a stray September poll into the 2017 figures? This guy. (changes vs Q4 2017);

  • SNP – 37.8% (+0.5)
  • Conservative – 26.8% (+1.9)
  • Labour – 25% (-3.3)
  • Lib Dem – 6.6% (-0.4)
  • Green – 1.4% (+1.4)
  • UKIP – 1.2% (+1.2)

At this point Labour have gone back  even from where they were at the snap election last year, which given the margins at play in many constituencies may concern them, whilst the Conservatives have bounced back a bit. The Greens and UKIP registered this time simply because only one poll didn’t prompt for them. But overall, this isn’t really a lot of movement, and it roughly establishes a post-snap status quo.

Again going via Electoral Calculus (changes vs Q4 2017);

  • SNP – 40 (+9)
  • Conservative – 12 (nc)
  • Lib Dem – 4 (nc)
  • Labour – 3 (-9)

What a difference those small swings makes to this projection; Labour knocked down to 3 seats, fewer than the Lib Dems would be expected to win on not much more than a quarter of Labour’s vote. But hey, FPTP, it gives the people what they want, right? Although the Conservatives are therefore back into second place in both votes and projected seats, the main beneficiaries of a Labour slump would unsurprisingly be the SNP.

There are also five polls for Holyrood, though note three of those are Panelbase, who’ve tended towards particularly high SNP and low Green and Lib Dem figures. The constituency vote averages out as (changes vs Q4 2017);

  • SNP – 40.4% (+0.4)
  • Conservative – 25.6% (+1.3)
  • Labour – 23.2% (-2.8)
  • Lib Dem – 7.2% (+0.5)
  • Green – 2.0% (+1.3)

And the list at;

  • SNP – 34.8% (+1.5)
  • Conservative – 24.4% (+2.2)
  • Labour – 22.0% (-2.3)
  • Green – 7.6% (-0.4)
  • Lib Dem – 7.6% (-0.4)
  • UKIP – 2.0% (-0.3)
  • SSP – 0.0% (-1.0)

As with Westminster, what seemed like a small Labour comeback at the end of 2017 has reversed as the Conservatives claw back up to 2nd place. The SNP are also up marginally, whilst it was much of a muchness for the two smaller parties, both registering small increases in constituency but declines in list support.

And in terms of seats (changes vs Q4 2017);

  • SNP 54 (nc)
  • Conservative – 33 (+5)
  • Labour – 27 (-4)
  • Green – 8 (nc)
  • Lib Dem – 7 (-1)

That means no change in the overall balance in terms of constitutional position, but within the pro-Union parties the balance shifts back in the Conservatives favour. The SNP and Greens are unchanged overall compared to this time last year, though the SNP would have picked up a couple of extra constituencies.

You may wonder how the Greens have the Lib Dems beat in seats despite identical vote shares meaning identical seats last time. Remember that the list vote is proportional within each region, not nationally. If you compare the two projection graphics side by side you can clearly see how this came about. Green MSPs are generally being elected one seat later, but none were the last seat elected on the Q4 2017 projection, so that doesn’t matter. However, the Lib Dems were projected to win the last seat in West, and that 0.4% drop combined with a Conservative increase was just enough to knock them out of it.

With all eyes on Brexit and the next Holyrood election looking quite distant, I wouldn’t expect this lack of polling drama to continue. As 2021 approaches it’s almost guaranteed we’ll see a lot more movement, but for 2017 vs 2018 it all proved rather quiet on this front, if not on any other!