GE19 – The Ballot Box Scotland Prediction

Polling day is finally here, and what would an election be without fevered speculation as to the outcome? Ballot Box Scotland isn’t immune to this particular fever, and as one final piece of pre-election coverage, I’ve come up with a list of 17 seats I think have the greatest potential to change hands. There are three big caveats that come with this. I am;

  • NOT Predicting a specific number of seats
  • NOT Predicting which party will win each seat on this list
  • NOT Predicting all other seats are safe and definitely won’t switch

No electoral system comes with that certainty, least of all the non-proportional and undemocratic nightmare that is First Past the Post.

Instead, I’ve given a range in which I think each party’s results may fall within. Note that the range isn’t as broad as “party loses all seats it holds that are down as possible flips” to “party holds all seats and gains all seats it is down as in contention for”. That’s because I think it’s very unlikely that, for example, the SNP would gain absolutely every Con-SNP marginal on the list, even though I do think they could win any of them individually. I’m not John Curtice however so this is based on gut-feel combined with a simple look at polling and past results, not rigorous statistical analysis of every seat.

Some of the seats on this list were less marginal in 2017 than seats that aren’t on the list, just because I don’t think they are likely to change hands. Argyll and Bute for example was closer between the SNP and Conservatives than Angus was, but since the Lib Dems are likely to do well there thanks to their bounce it strikes me as less likely to switch hands. Similarly, Labour’s very poor polling means I’ve simply written off the prospect of them gaining any seats, despite many close SNP-Lab marginals existing.

Final Polls

As we trot off to to vote today, here’s how things were looking according to the average from the last pre-election poll by Panelbase (3rd-6th), YouGov (4th-10th) and Survation (10th-11th). Note that Survation might still be a bit off because of some absolute nonsense that was “not publishing Scottish data til 6am” and then 6am coming and going and them… not publishing Scottish data. If it is wrong it’s the Courier’s fault. 

  • SNP – 41.0% (+4.1)
  • Conservative – 28.0% (-0.6)
  • Labour – 20.3% (-6.8)
  • Lib Dem – 9.0% (+2.2)
  • Green – 0.8% (+0.6)
  • Brexit – 0.8% (+0.6 vs UKIP)

Overall the SNP are up reasonably since 2017, the Lib Dem bounce is present but weaker than expected, the Conservatives down so little we might as well call it static, and Labour facing the prospect of their worst ever UKGE vote share. In the briefest possible terms if these figures are near correct then all of Labour’s seats bar Edinburgh South are at serious risk, the Conservative performance may rely on them having concentrated their vote share further into the seats they currently have plus some key further targets, and the Lib Dems still aren’t in the running for gains anywhere except North East Fife.

Underneath that average, there’s quite a lot of variance within those polls. For example Panelbase found the SNP to be on 39%, whereas other agencies reckoned they were in the 40s. Likewise, Survation found a dramatically low 7% for the Lib Dems which is no improvement on 2017, versus 10% for the other two.

Some folk will look at past elections and say “well, they under/overestimated X party last time…” but that absolutely does not mean the same thing will happen this time. We won’t know til we have all the results in tomorrow, so we’ll have to make do with this as our most informed guess for now.

Seat Estimates

  • SNP – 38 to 46
  • Conservative – 8 to 15
  • Labour – 1 to 3
  • Lib Dem – 4 to 5

Presentation format; Constituency (Marginal rank, majority in votes, majority in percentage)

Close SNP-Conservative Marginals

  • SNP held
    • Perth and North Perthshire (2nd, 21, 0.04%)
    • Lanark and Hamilton East (8th, 266, 0.5%)
    • Central Ayrshire (17th, 1267, 2.8%)

Although the SNP are up in the polls, these are seats where the margin between them and the Conservatives is so tight that it’s perfectly possible to imagine them changing just as part of the general churn in votes, especially if the Conservatives have campaigned well in the area. Lanark and Hamilton East is particularly worth a watch as it’s notionally a three-way marginal based on the 2017 result, but I really can’t see Labour being in the running for it anymore.

  • Conservative held
    • Stirling (5th, 23, 0.3%)
    • Gordon (22nd, 2607, 4.9%)
    • Ayr, Carrick and Cumnock (28th, 2774, 6.0%)
    • Ochil and South Perthshire (30th, 3359, 6.2%)
    • Angus (32nd, 2645, 6.6%)

All of these seats have margins either within or only slightly above tipping to the SNP on a uniform national swing. That is of course not how elections work in reality, and all five could remain in Conservative hands if their vote has concentrated within them. But none of these would be a surprise if they switched.

Outside SNP-Conservative Chances

  • East Renfrewshire (42nd, 4712, 8.8%)
  • Aberdeen South (48th, 4752, 10.7%)

Both of these Conservative held seats are quite safe by current Scottish standards on paper, but have unique dynamics that could undermine Conservative chances. East Renfrewshire saw the second highest Remain vote in Scotland in 2016, making it one of the most pro-EU places in the UK. It could be somewhere that 2017 Conservative voters unsettled by Boris Johnson’s approach to Brexit and likewise unwilling to support a pro-Independence party pile behind the Lib Dems. That could help the SNP’s chances.

Meanwhile former Aberdeen South MP Ross Thomson didn’t seek re-election following sexual harassment allegations. Combined with an antisemitism, Islamophobia and homophobia scandal for the party’s now-suspended candidate in neighbouring Aberdeen North, and if there’s a city wide distaste for Conservatives, that again plays in the SNP’s favour.

Labour-SNP Marginals

  • Rutherglen and Hamilton West (7th, 265, 0.5%)
  • Kirkcaldy and Cowdenbeath (9th, 259, 0.6%) NOTE: Candidate nominated by SNP has since been disclaimed by the party
  • Glasgow North East (11th, 242, 0.8%)
  • Midlothian (14th, 885, 2.0%)
  • Coatbridge, Chryston and Bellshill (30th, 1586, 3.5%)

The bulk of Labour’s MPs are at serious risk of losing their seats to the SNP. Solid local campaigns can protect against a national tide, so I wouldn’t be entirely surprised if any of these individually stay red, but I’d be utterly astonished if many did given the likelihood of a result below even their 2015 calamity.

Their best chance of these is likely in Kirkcaldy and Cowdenbeath, where Neale Hanvey has been suspended from the SNP and had support for his campaign withdrawn over issues with antisemitism. He could still be elected, which technically makes this a Lab-Ind marginal, but if enough voters are aware of and (rightly) put off by that scandal, Lesley Laird will be re-elected.

Messy Lab-SNP-Con Contest

  • East Lothian (25th, 3083, 5.5% vs SNP, 3618, 6.5% vs Conservative)

By no means the only three way marginal in the country (we mentioned Lanark and Hamilton East earlier), it is the only one I think is genuinely in play for all three in this election. Labour will be facing off against both a resurgent SNP and a Conservative party that’s definitely been targeting the area. Could go any way.

Help Me, Fifey-Wan Kenobi, You’re My Only Prospective Gain

  • North East Fife (1st, 2, 0.005%)

North East Fife was literally the most marginal seat in the UK never mind Scotland in 2017, and both parties have sunk vast amounts into this one. Anyone telling you they are certain which way this one will go is an absolute charlatan. It is however the only seat the Lib Dems have any serious prospect of gaining in Scotland at this election by basically anyone’s reckoning.

I actually think that the Lib Dems COULD be at risk of losing one or two of their other seats on the basis of the Survation poll, but thanks to it being so unbelievably and frankly unacceptably last minute I’m sticking to what I had settled on at 1am this morning because I simply do not have the time to make any further changes. Effectively although they may look shaky in a couple of places on the basis of uniform swing, more than any other party the Lib Dems are about concentrated vote shares. They’ve gained constituencies and lost votes in both 2016 SP and the 2017 GE, so expect them to hold on to what they have if they are up.

Also I know the header is a REALLY weak joke, wheesht.

Smaller Parties

Smaller parties aren’t in the running for any seats, and First Past the Post depresses their prospective vote share anyway. Instead, I’m thinking about where they might achieve at least 5%.

  • Greens – 1 to 5 held deposits
    • Glasgow North, possibly other Glasgow and Edinburgh seats, perhaps Kirkcaldy and Cowdenbeath depending on how many Pro-Independence but anti-Hanvey voters switch.
  • Brexit – No held deposits
    • They aren’t standing in any of the most pro-Brexit seats because those are Conservative held, and Conservatives squeezing their vote anyway.
  • UKIP – No held deposits
    • There’s no overlap between UKIP and Brexit candidates so they’ll likely pick up the hard line Pro-Brexit vote where they are standing, but even in Moray don’t expect much.