As the General Election continues to approach at breakneck speed, the eyes of the UK’s media are firmly on the prize of Downing Street and the Conservative Party’s desperate fight to avoid the worst defeat in their history. Up here in Scotland however, a Conservative vs Labour narrative feels almost like a sideshow to the real contest between the SNP and Labour.
After a solid decade of defeats, including all time lows in both the 2019 UK and 2021 Scottish elections, Scottish Labour have a spring in their step last seen as they (with hindsight, overconfidently) approached the 2011 Scottish election following their big 2010 wins here. Although consistent growth in the polls started with the Conservative collapse throughout 2022, since the sudden resignation of Nicola Sturgeon as SNP Leader and First Minister last spring, they too have been leaking support to Labour.
According to the Ballot Box Scotland trackers for both the upcoming General Election and the Scottish Parliament, Labour have pulled ahead of the SNP for the first time since the 2014 Independence Referendum. Avid followers of BBS will be aware that as Labour began to recover in the polls, I’d identified a 10% lead for the SNP as the beginning of a “danger zone” for them, where they would begin to lose increasing numbers of seats to Labour. As that gap narrowed and then began to open back up but now in Labour’s favour, the expected seat losses have naturally gotten worse and worse for the SNP.
Projections from recent polls have suggested they could lose two-thirds or more of their seats, whilst Labour could add dozens to the single seat they won in 2019. We’re all used to First Past the Post giving exaggerated results, which is part of why the system is so democratically rotten, though for the past three elections the system has very much been to the SNP’s rather than Labour’s advantage. Even so, some of these numbers have felt overly dramatic given the relatively narrow expected lead in votes.
The state of play
To figure out exactly what’s going on here, let’s start with where things currently stand per the BBS GE24 average (available on the hub page here).
Polling Average (28th of May)
(Note that on the GE24 Hub page, the Greens and Reform UK are listed as 2.8%, due to two polls with no separate figures for those parties counting as 0% in the average. Since they wouldn’t be on 0% in reality, the seats chart shows a proportional ideal, and they were just off the 3% threshold I use, I gave both 2% instead of 0% in those polls purely for illustrative purposes!)
Seat Projection from Average (28th of May)
Although Labour’s current polling average is a clear-but-modest 5.6% over the SNP, both the BBS and Electoral Calculus models translate that into winning just over double the number of MPs. We can see just how completely out of kilter FPTP is with actual vote shares through the proportional ideal bar on the bottom. Rather than the 15-16 seat lead over the SNP both FPTP models give, a fair, proportional system would give them a mere 3 seat advantage.
Their projected seat tally is especially remarkable when you consider that in 2019 an 11.5% lead for the Conservatives over Labour at UK level, and 43.2% of the vote overall, “only” (all things are relative) gave them 1.8x as many seats. So how are Scottish Labour potentially turning the support of just over a third of the electorate and a modest lead into a clear majority of MPs and a thorough drubbing for the SNP?
Voters: It's where they are, not how many you have
I mean, we all know that FPTP is about how your vote is spread, that isn’t news to anyone. However, the specifics of that basic fact of FPTP life are often misunderstood in Scotland. Many people recall a time when Scotland was a “Labour country”, but that never really existed. Even when Labour were at the absolute peak of their success, they weren’t winning MPs in places like the Borders, Perthshire, Angus, Aberdeenshire, Moray, or Argyll. In elections as in so much else in Scotland, popular discourse has tended to elide the Central Belt, home to the majority of the population, with the nation, and in so doing give an incomplete picture.
What Labour were doing in their heyday was winning absolutely staggering, unchallengeable majorities in the urban Central Belt, whilst coming in as hopeless also-rans in most of the rest of the country (notable exceptions including Na h-Eileanan an Iar or Dumfries and Galloway). That’s also partly why the SNP’s success in raw seat terms has been so much greater; it hasn’t simply been down to the size of their vote, but the fact it has been a nationwide party in a way Labour never have been. What is an advantage with a high national vote share however becomes a serious problem as your share drops, allowing more geographically concentred parties to make big gains at your expense.
Categorising Constituencies
To see the effect this has had historically, I’ve divided every constituency in the country up into five categories. The biggest category, with 33 seats on the new boundaries (34 on the old set) is Central Belt/Urban, which is obviously most of the Central Belt, plus southern Fife (controversy abounds as to whether that is Central Belt), Aberdeen and Dundee. These are complimented by 6 Urban Affluent seats, which capture some of the bits of urban Scotland with the greatest concentration of affluence; in political terms, these are the urban seats with a recent-ish history of non-SNP or Labour representation, plus Edinburgh South, which has an unarguably posh core.
The rest of the country is then split between contiguous geographic blocs. South contains 4 seats, covering the Borders, Dumfries and Galloway, southern Ayrshire, plus most of Clydesdale in South Lanarkshire. Mid/North East is 8 seats, basically North East Fife, Stirling, Perth and Kinross, Aberdeenshire, Angus and Broughty Ferry from Dundee. Lastly, Highlands/Islands with 6 seats (7 under the old boundaries) is self explanatory: Highland, Argyll and Bute, Na h-Eileanan and Iar, Orkney and Shetland.
These are relatively simplistic categorisations just to illustrate the point. Within the “Mid/North East” grouping for example there are significant distinctions between areas with a strong Lib Dem history (North East Fife, much of Aberdeenshire) versus those where the pre-referendum scraps were between the SNP and Conservatives (Perthshire, Angus, north Aberdeenshire). You could do further breakdowns on that basis for a more nuanced picture, but since Labour are our primary focus here, that’s not as necessary.
UK General Election 2019 votes by constituency category
(Note that these are the categorisations on the previous boundaries, so that I could compare across multiple years, whereas the earlier map was the categorisation on the new boundaries. This has very, very little impact on the distribution of votes between categories.)
This demonstrates exactly that Labour spread of votes I was talking about earlier: note how their vote share in Central Belt/Urban seats is almost 7% better than their national share, but they do worse in every other category, bottoming out at 13% worse than nationally across the Mid/North East seats. The SNP by contrast only exceed their national share in Urban/Central Belt seats by 2%, albeit that is still their strongest segment, whilst only coming about 6% lower in the Urban Affluent seats.
SNP vs Labour margin by constituency category
If we focus in on the contest between the SNP and Labour, and extend it back over the past two elections, we can see there’s a consistent pattern of the margin between the SNP and Labour being narrower in both urban categories….
SNP vs Labour Margin by Constituency Category (Relative to National)
Another way to visualise the same data would be to show how much (/less) closer Labour’s vote share was to the SNP’s in each category, relative to the national margin. A negative value here means the SNP vs Labour margin was narrower than nationally, and thus a positive value means a wider margin. Straight away you can see how Labour melted away after 2015 in the South (2 of 4 MPs in 2010 to 0 runner ups in 2017) and the Urban Affluent (4 of 6 MPs in 2010 to 1 MP and 0 runner ups in 2017) seat blocs, replaced largely by the Conservatives.
Yet at the same time, you can see how from 2015 to 2017 they managed to shore up their position in Central Belt/Urban seats, though they lost ground to the Conservatives in the Urban Affluent bloc. Then, despite their collapse in overall vote share in 2019, they managed to preserve a roughly 4-5% narrower margin in those constituencies than they had nationally. What this means in the simplest of terms is that as Labour’s vote share increases, we can expect them to overtake the SNP in these seats before it does nationally.
Blood Red Summer
All of this has huge implications given that in 2019 the Central Belt/Urban and Urban Affluent constituencies accounted for 69% of the votes cast in Scotland, and will be 39 of the seats up for grabs this time. Even if the two are tied in votes nationally, Labour can be expected to have the advantage in the clear majority of seats nationwide. Hence why that relatively modest 5.6% lead for Labour at the moment is nonetheless projected to leave barely a splash of yellow across the map in the Central Belt.
Indeed, if we map Labour’s notional vote shares in 2019, we can get a pretty good idea of what parts of the map will look like in July. Apart from their strong starting point in Na h-Eileanan an Iar, you can see the deepest reds running almost unbroken from coast to coast across the Central Belt. East Lothian, Midlothian, North Lanarkshire and Glasgow in particular already have Labour starting with a significant advantage relative to their national performance, and it’s highly likely that has positioned them to pick up every seat covering those areas.
One final note of caution for anyone about to get really carried away looking at the “Swing Required” data on this map: the entire point of this piece has been that Labour can expect their strongest results in the urban Central Belt. Whilst we can certainly expect to see growth nationwide for them, don’t make the mistake of trying to impose a national swing reading of the polls onto the likes of Gordon and Buchan or Berwickshire, Roxburgh and Selkirk.
All of this said, voters are nothing if not unpredictable, and like to do all sorts of weird and wonderful things. Amongst the expected SNP losses, there may be MPs who manage to hold out against a national tide in the same way Jackie Baillie has for Labour in the Scottish Parliament. At the same time, some of the seats the SNP feel safer in could unexpectedly slip from their grasp. Add in the fact we still have over four weeks of this campaign left to go, and there’s always room for surprises. For now though, after a long period of suffering from the deficiencies of FPTP, it looks like it’s about to give Scottish Labour something to celebrate for the first time in 14 years.
If you find this or other Ballot Box Scotland output useful and/or interesting, and you can afford to do so, please consider donating to support my work. I love doing this, but it’s a one-man project and takes a lot of time and effort. All donations, no matter how small, are greatly appreciated and extremely helpful.
(About Donations)