There are a number of things that the 2021 Holyrood election are about. The first is one of whether pro-indy or anti-indy parties win the overall majority. As things stand right now, I simply do not see a path for pro-indy parties to lose their majority - There are 59 constituencies that the SNP are prohibitive favourites in, and at least 5 Green list seats that are (in my opinion) very safe. That takes them right up to the line, before factoring in further seats which the SNP and Greens are also likely to win. Therefore, barring something truly earth-shattering the main play for unionist parties would be to delegitimise a combined SNP-Green majority as not being a valid mandate for pursuing independence. This of course, lends itself to the running SNP message of “only a majority for us can legitimise indyref2”, which then becomes the second dominant question of this election: whether the SNP becomes the majority party yet again after narrowly losing it in 2016.
Given the vagaries of D’Hondt, a list system designed to counteract FPTP dominance, the SNP’s simplest and most watertight path to a majority is to win at least 65 out of 73 constituency seats. In order to deny the SNP a majority off of constituencies alone, they need to keep at least 8 seats - as things stand, there are only four constituencies where unionist parties are the prohibitive favourite - Orkney, Shetland, Ettrick and N.E. Fife. There are another four constituencies where unionists are leaning towards being slight favourites - Edinburgh Western, Edinburgh Southern, West Galloway, and Dumfriesshire. Although a pro-indy majority is all but guaranteed from the current state of the parties, an SNP majority is right on the line.
Most models and election hacks treat the question of constituencies as one of pure swing alone, as though looking at national voteshare changes and applying them uniformly to all 73 seats will result in a faithful-ish reproduction of what the outcome of the election would look like on those percentages. Unfortunately, I don’t think this is really sufficient to understand the likely outcome of the election. Universal swing may well have held in a political environment where a certain level of depolarisation and ticket-splitting (e.g. voters who used to opt for one party at Westminster and another at Holyrood) prevailed, but increasing partisan identification and Yes/No polarisation has greatly reduced the salience of split-ticket voting. And the EU referendum permanently scrambled the geographic and social coalitions of the parties in ways that would be completely unrecognisable to the voters of 2016, when the last Holyrood election was held. For that reason, my analysis of constituencies takes into account the voting patterns for the coterminous seats that have gone to general elections twice (2017, 2019) since the last devolved election.
On that basis, the most straightforward path to an SNP majority goes through
(A) Defending the vulnerable seats they currently hold, including ones which have now voted Conservative twice at the Westminster level.
(B) Flipping seats held by Unionists which the SNP already nominally hold or gained at the Westminster level.
The SNP hold 59 constituencies at present. By my count, there are four seats which the SNP need to be vigorously defending - let’s call this the “Protect” column. There are then six seats where they ought to be on the offensive, the “Must Win” column. Of the latter, four are seats which voted SNP in 2019, and another two are seats which have trended significantly towards the party. There are another four seats which are *theoretically* very gettable and could well bolster the SNP’s seat count, but where trends since 2016 and the stubbornness of incumbency make them less viable targets. Call it the “Would Be Nice” column. In the following weeks, I intend to go through all of them here. Of course, the twists and turns of the election campaign may make some of my assumptions irrelevant, which will be reflected here faithfully.
PROTECT:
Moray (S)
Banffshire & Buchan Coast (S)
Aberdeenshire East (S)
Perthshire South & Kinrossshire (S)
MUST WIN:
Eastwood (C)
Edinburgh Central (C)
East Lothian (L)
Dumbarton (L)
Ayr (C)
Aberdeenshire West (C)
WOULD BE NICE:
Edinburgh Southern (L)
Edinburgh Western (LD)
Galloway and West Dumfries (C)
Dumfriesshire (C)