The battle for the Scottish marginals

It is a truth universally acknowledged that at every general election, most seats can be ignored when trying to predict the outcome. The real battle is only in a sub-set of seats: the marginals. And at this election, a lot of these are in Scotland.

What happens in Scotland is important not only in the sphere of Scottish politics and the continuing debate around Indyref2, but also in relation to the outcome for the UK as a whole. So where are these seats and how might they shape the overall result? 

Marginals are normally defined as seats where a swing of up to 10 per cent will lead to a change. In the 17 elections since 1951, there has only been one election, Tony Blair’s first victory in 1997, where more than a handful of seats have seen a larger swing than this between the two major parties. The average Con/Lab swing in the last 70 years is in the region of 3 per cent.

At this election, around a quarter of seats across the whole of Britain are marginals: 169 of 650 seats. In contrast, more than three quarters of Scottish seats are marginal: 46 of 59. Come election night, lots of the attention will be on the results in Scotland. 

Thirty of the Scottish marginals are currently held by the SNP. If there was a swing away from the SNP, these would be the main battleground. However, recent polling, such as last week’s Ipsos MORI Scotland poll, has pointed to a rise in the SNP’s share of the vote. So most of the interesting marginals will be in non-SNP held seats. Let’s focus on them. In all of these marginals, the SNP came second in 2017. 

Labour currently holds seven seats. Of these, six are marginal. Indeed, the party’s second safest seat, East Lothian, requires a swing of only 5.5 per cent for the SNP to take it. If it achieves this, Ian Murray in Edinburgh South will probably be the only Labour MP left in Scotland. Again.

For the Lib Dems, one of their four Scottish seats is marginal, Edinburgh West. It would take a swing of 5.7 per cent to the SNP to unseat them here. The Lib Dems will hope that this seat doesn’t turn into a tight contest. If it does, it will signal that the fortunes of Jo Swinson’s Lib Dems are going backwards in her home country.

However, most interest will be in the Scottish Tory seats—and if their good performance in 2017 can be replicated without Ruth Davidson as their leader. With polls suggesting gains for the Tories in England and Wales but that their support in Scotland is being out-stripped by the SNP, the Conservatives are facing the prospect of victories south of the border being partially offset by losses north of it. This makes what happens in Scotland important in terms of whether the Tories will get a workable majority. 

The Conservatives currently hold 13 seats in Scotland. In all of these, the SNP came second in 2017. Should the swing to the SNP from the Tories amount to less than 4.9 per cent, the Tories are in danger in only one, former Scotland Secretary Michael Forsyth’s old seat of Stirling.

However, any more than a 4.9 per cent swing, and very small increments could make a big difference in seats. A swing of between 4.9 per cent and 6.6 per cent, and up to a further four seats could switch to the SNP. On a swing of between 8.7 per cent and 10.9 per cent, up to five more would go. As such, the difference between the Tories holding 12 of their 13 seats in Scotland, or just three of them, is very small in terms of share of the vote.

Of course, should the Tories outperform the current polls and enjoy an upsurge that outstrips that of the SNP, there are five seats where a very small pro-Conservative swing would see the Tories gain from the Scottish nationalists: Perth and North Perthshire, Lanark and Hamilton East, Edinburgh South West, Argyll and Bute, and Central Ayrshire.

For the Lib Dems, only in North East Fife does it look like they have any chance of a gain in Scotland. This is the most marginal seat across the whole of the UK—only two votes separated them from the SNP in 2017.

So, seven seats to watch….

- Tory v SNP—Perth and North Perthshire (swing needed, 0.04 per cent to the Tories): The Conservatives’ number one target in Scotland. Win here, and they should be increasing their representation in Scotland.

- SNP v Tory—Ayr, Carrick and Cumnock (swing needed, 6 per cent to the SNP): Gain this, and the SNP would expect to take at least three seats from the Tories.

- SNP v Tory—Moray (swing needed, 8.7 per cent to SNP): If the SNP wins here and the level of swing is reflected throughout Scotland, the SNP will gain at least six seats from the Tories.

- SNP v Tory—Dumfries and Galloway (swing needed, 10.9 per cent to SNP): If the SNP takes this seat and it is reflective of the national trend, it’s very likely that it will gain at least 10 of the 13 Tory seats in Scotland.

- LD v SNP: North East Fife (swing needed, <0.01 per cent to the Lib Dems): The most marginal seat in Britain. With a current majority of two, this is one seat where no one can quibble that a Lib Dem bar chart is overstating the closeness of the race.

- SNP v LD: Edinburgh West (swing needed, 5.7 per cent to the SNP): If the SNP wins here, it’s likely to make gains from all three other major parties and come close to repeating its remarkable performance in 2015.

- SNP v Labour—East Lothian (swing needed, 5.5 per cent to SNP): If the SNP gains this, Labour is likely to lose all bar Ian Murray’s Edinburgh South to the SNP.

One final point to add to the margins. These calculations are all based on Uniform National Swing. While there have always been seats that buck the national trend, this method has been pretty good historically at accurately predicting how the share of the vote will translate into seats. However, some have questioned whether UNS will be as accurate in this election as it has been in the past. For methodologists, one fascinating aspect of the election will be how well UNS performs against the newer alternative way of estimating seats from share of the vote: MRP or “Multilevel regression post-stratification.”

Either way, on election night considerable attention will be focused on results north of the border.