In the previous post we looked at the swings across the country that helped transform the electoral map. Although the difference between 2015 and 2017 is a warning to all parties never to rest on large majority laurels, the overall effect of GE19 was to make Scotland a much less marginal country. We’ve gone from a whopping 46 marginal constituencies to just under half as many with a much more manageable 22. Whenever the next election rolls around, it now has a completely different set of key battlegrounds to the top 20 most marginal seats I covered in the run up to GE19.
Given that the SNP made substantial gains at the expense of both the Conservatives and Labour, and the Lib Dems are still very concentrated in a handful of seats, it’ll come as no surprise that most seats the SNP didn’t win are marginal. Last time, a neat third (8) of the 24 non-SNP seats weren’t counted as marginals, though the two lowest majorities out of those still flipped. This time, it’s 9 out of the 11 are marginal.
That isn’t the same as saying the SNP are going to take many or any of them at the next election, as who knows what will happen between now and then. It just means that in relative terms, the other parties have a lot more to lose if the next election doesn’t go their way than the SNP do. Should the SNP do poorly next time the Conservatives are best placed to capitalise, as out of the other 13 marginal seats, they are the runners up in 8 of them.
Let’s take a quick look at how seats moved up and down the marginality rankings. This is effectively safety in relative terms. Green seats moved towards the safer end of the scale, and red seats moved towards the less safe end.
Overall, seats the SNP won in 2017 and those they won from Labour this time moved up towards the safer end of the pack, whilst most non-SNP seats in 2017 went in the other direction. That’s not universally true however. The one seat they lost, North East Fife, is relatively more safe than last time as it ceased to be the number 1 marginal in Scotland (and the whole UK). Stirling was so marginal last time and such an emphatic SNP gain that it also moved well into the safer half of the rankings.
Looking at the best and worst performances, West Aberdeenshire and Kincardine experienced the most dizzying drop. It plummeted 49 places from the 54th most marginal (i.e. the 6th safest) seat in the country with a Con over SNP majority of 15.4% to the 5th most marginal with a majority of just 1.6%. East Dunbartonshire, which of course did change hands and deprive the Lib Dem leader of her seat, fell 46 places to go from the 47th most marginal with a LD over SNP majority of 10.3% to the most marginal seat in the country at 0.3% SNP over LD.
Going the other way, Edinburgh South West leapt up 32 places from 15th most marginal and an SNP over Con majority of 2.2% to 47th most marginal at SNP over Con of 23.0%. Stirling went from the 5th most marginal and a very slender Con over SNP majority of 0.3% up 30 places to the 35th most marginal and a lead of 17.6% SNP over Con.
That tells us which direction seats moved in the rankings but it doesn’t actually tell us whether a given seat became more or less secure in absolute terms. It’s possible for example for a seat to move down the rankings because the majority there increased by less than somewhere else. Green seats saw an increase in the absolute size of the winner’s majority (i.e. just the size of it, doesn’t account for who won each time), whilst red saw a narrower majority.
Much more green on this map. That shouldn’t come as a surprise given we know there are half as many marginals now as last time, so a lot of seats therefore have to have increased their majority. The overwhelming majority of seats, 47 of them, now have a safer winning margin for the party that holds them now than the 2017 winner did. Only 12 are less safe than in 2017. Remarkably, that number only includes 4 of the 14 seats the SNP gained. In most seats they gained, they set a majority larger than the one they overturned to take the seat. Their majority also increased in every seat they successfully held.
That means that 8 of the 12 seats that are less safe than last time are in the hands of the Conservatives, Labour and Lib Dems. Of the seats they held going into the election, only Edinburgh West (Lib Dem) and Banff and Buchan (Con) saw their incumbents increase their majority. The one gain from the SNP in North East Fife (Lib Dem) also has a larger majority than last election, as noted earlier, just by virtue of there being more than 2 votes in it this time.
Putting that all together, we’ve got a new list of key battlegrounds for the next election. As there were so many last time I only focused on 20 of them – since there are only 22 marginals at all now, I can more easily list them all. Seats that were part of the top 20 most marginal covered in Ballot Box Battlegrounds are highlighted in bold.
- East Dunbartonshire, SNP over LD 0.3% (was 47th, LD over SNP 10.3%)
- Caithness, Sutherland and Easter Ross, LD over SNP 0.6% (was 34th, LD over SNP 6.6%)
- Moray, Con over SNP 1.1% (was 41st, Con over SNP 8.7%)
- Gordon, SNP over Con 1.5% (was 22nd, Con over SNP 4.9%)
- West Aberdeenshire and Kincardine, Con over SNP 1.6% (was 54th, Con over SNP 15.4%)
- Kirkcaldy and Cowdenbeath, SNP over Lab 2.6% (was 9th, Lab over SNP 0.5%)
- North East Fife, LD over SNP 2.9% (was 1st, SNP over LD 0.005%)
- Dumfries and Galloway, Con over SNP 3.5% (was 49th, Con over SNP 10.9%)
- Ayr, Carrick and Cumnock, SNP over Con 5.0% (was 28th, Con over SNP 6.0%)
- East Lothian, SNP over Lab 6.6% (was 25th, Lab over SNP 5.5%)
- Edinburgh West, LD over SNP 6.9% (was 39th, LD over SNP 5.7%)
- Glasgow North East, SNP over Lab 7.5% (was 11th, Lab over SNP 0.8%)
- Dumfriesshire, Clydesdale and Tweeddale, Con over SNP 7.7% (was 56th, Con over SNP 19.3%)
- Ochil and South Perthshire, SNP over Con 7.8% (was 30th, Con over SNP 6.2%)
- Argyll and Bute, SNP over Con 8.6% (was 16th, SNP over Con 2.8%)
- Aberdeen South, SNP over Con 8.7% (was 48th, Con over SNP 10.7%)
- Angus, SNP over Con 8.8% (was 32nd, Con over SNP 6.6%)
- Berwickshire, Roxburgh and Selkirk, Con over SNP 9.7% (was 58th, Con over SNP 21.1%)
- Rutherglen and Hamilton West, SNP over Lab 9.7% (was 7th, Lab over SNP 0.5%)
- Banff and Buchan, Con over SNP 9.7% (was 43rd, Con over SNP 8.9%)
- Lanark and Hamilton East, SNP over Con 9.8% (was 8th, SNP over Con 0.5%)
- East Renfrewshire, SNP over Con 9.8% (was 42nd, Con over SNP 8.8%)
With a substantial Conservative majority in the Commons now overall, it’s unlikely I’ll need to come back to this list to run another Battlegrounds series until at least 2023. However, you can expect a Scottish Parliament equivalent ahead of 2021.