As the 2026 Scottish Parliament election approaches, keep up to date with all the latest polling and analysis on Ballot Box Scotland’s Holyrood Hub page!
Pity the poor Scottish Labour Party! Two years ago, they were screaming and punching the air in celebration as they steamrollered constituencies across the Central Belt in a stunning first victory in over a decade. Now, they are desperately talking themselves into believing that polling has got it wrong or that tactical voting can still get Anas Sarwar into Bute House.
This was not how the first UK Labour Government of my voting lifetime was expected to go. Anas Sarwar closely aligned himself with Starmer’s vision, having managed to coup the Corbyn-era and aligned Richard Leonard mere months before the 2021 election. Prime Minister Starmer was, from the Scottish perspective, to be a step towards First Minister Sarwar.
Yet beset by a great many troubles, rather a lot of them of its own making, the deep unpopularity of that government has all but killed the hopes of their Scottish and Welsh arms. Sarwar’s big gamble of calling for Starmer to resign seemed instead to bolster the latter’s position: in the same day, his Premiership went from seeming like its remaining time could be measured in hours, to clearly at the helm up until what the party has acknowledged will be a bloodbath set of votes in May.
Barring the biggest polling miss in Holyrood’s history, Scottish Labour are no longer meaningfully in the race for government: they simply need to survive.
Holyrood History
Note: For consistency, all current polling average references and projections in the Party Profiles series are based on the last data point by the 1st of April 2026, as it was at time of publication.
Labour Vote Share at Scottish Parliament Elections
Labour have the unenviable distinction of being the only major Holyrood party to have lost support, in both votes, at every election in the parliament’s history. You could argue that 1999 itself meets that criteria, as it used identical constituency boundaries to Westminster (barring the separation of Orkney and Shetland) at that time, yet Labour’s constituency vote was 7% shy of what they won in 1997. Nonetheless, they easily formed the first Scottish Executive (as the early Scottish Governments were called) with the Lib Dems, and held onto it in 2003.
When they lost government for the first time in 2007 it wasn’t really due to a loss of their own support, which barely changed, but instead an increase in the SNP’s. Indeed, the same can be said of what at the time felt like an absolutely crushing loss of stronghold constituencies in 2011, when their constituency vote wasn’t even down a whole half a percentage point. As we know with the benefit of hindsight 2011 was mild compared to what followed the referendum, as the party crashed into third place in 2016 and stayed stuck there in 2021.
The current polling average suggests that will continue, as the party sinks below 20% on both votes for the first time. They’d already gone into the teens on the list in 2016, but their constituency vote falling so far is new. In an election where the SNP are expected to massively warp the proportionality, the constituency vote matters more than ever, and they can ill afford this. One tiny, tiny silver lining for Labour has been that at the time of writing, their polling average on that vote is now tied with Reform, rather than behind. It remains within the margin of error and thus an effective statistical tie, but there’s nonetheless a psychological boost to be had from that.
Labour Seats Won at Scottish Parliament Elections
A frustratingly common myth about the Holyrood voting system is that it was “designed to stop an SNP majority.” This isn’t true, or rather, it combines two truths into an untruth. The Scottish Parliament voting system is meant to be proportional: true, but that makes it hard for any party to win a majority, not just the SNP, and it was the Lib Dems who insisted on that. The exact form of PR, with a huge FPTP element, was accepted by Labour as they thought it would benefit them: true, but not only would it have been unthinkable that the SNP would eclipse them in the Central Belt, this equally negatively impacted the Conservatives.
Having said that Labour have lost votes at every election, naturally that translates to losing seats. Indeed having said you could argue 1999 started with a decline in votes, the same is true of seats. Despite 57 seats more being available than at Westminster, Labour’s 56 matched their 1997 constituency haul entirely, and even then it required 3 list seats to do so: relative to 1997, they lost a constituency to each of the SNP, Lib Dems and a formerly Labour Independent, Dennis Canavan.
Constituency losses continued over the next two elections, but when they lost to the SNP in 2007, Labour still won the barest majority of constituency seats overall with 37 of the 73. Much maligned by low-info Cybernats, it was the list system which was absolutely essential to giving the SNP the lead in seats it was due based on votes, and one that should have been slightly wider but for Labour still having too many constituency wins.
2011’s constituency losses weren’t shattering just because they represented many decades-old strongholds swinging to the SNP, they also meant the loss of most of their frontbench and experienced MSPs, as the party at that time had a truly stupid policy of barring dual candidacies. If you’re a long time fan of BBS you’ll have seen me make this reference a million times, but I’m doing it again: this was so stupid even New Zealand’s Electoral Commission pointed at it and said “well, that was stupid of you” (page 38).
By 2016, constituency representation was to be a tiny proportion of the Labour contingent at Holyrood. The system they accepted thinking it would benefit them had turned against them, as it had the year before at Westminster. For the tiniest mercy, by the time 2021 rolled around, the further losses they experienced could at least be described as the smallest in their history, losing just two seats. Things can still get worse though, as the BBS model currently puts them on just 18 seats in total. By contrast their best pre-GE24 polling put them on triple that number, 53 MSPs, nearly matching their 1999 haul.Â
Labour List Vote Distribution by Region in 2021 (Notional)
Perceptions of Scotland as a “Labour Country” from 1959 onwards have never really been true, instead reflecting the distorting effect of First Past the Post. Labour had such strong support across urban Scotland that their vote was weighed rather than counted, but they never had a comparable presence across much of rural Scotland, such as the Aberdeenshire, Angus, Argyll, (lots of A’s) the Borders, Perthshire or Moray.
That pattern has continued up to the present day, with their 2021 distribution only crossing 20% in three of the four Central Belt regions, only narrowly missing it in Edinburgh and Lothians East due to the presence of the only Lib Dem constituency in those regions. Mid and Fife and South were weaker as whilst they have areas of historic strength (southern Fife and Ayrshire, respectively), they’ve also got some of those traditional weak spots.
The North East is similar despite the presence of Scotland’s third and fourth largest cities as the rural component is larger. Highlands and Islands brings up the rear, and 2021 was the first time they fell below 10% and only elected a single MSP in any region in Holyrood’s history. On current polling, we might expect the North East to follow suit on at least the MSP count, if not possible the vote measure as well.
Key Areas
For brevity, I’ve picked out five of the most important places for Labour in this election. Not being included in this list doesn’t meant a constituency or region isn’t important, or that Labour aren’t in with a chance of winning it. You can find more detail about other interesting seats in the Ballot Box Battlegrounds series, and information about every constituency in region via the Regional Preview series.
Edinburgh Southern (Constituency)
Notional 2021 Constituency Vote
2026 Constituency Vote (Projected)
One of just two constituency seats Labour won in 2021, Edinburgh Southern has had substantial boundary changes that make it a genuinely southern seat, rather than its previous incarnation as effectively south-central. In theory this narrows their advantage over the SNP, but it’s important to note this makes the constituency a near perfect copy of the Westminster constituency. You know, the only seat Scottish Labour won in 2015 and 2019?
Now sure, the Holyrood and Westminster voting patterns are a little different, but are they that different? Especially with the benefit of another five years of the Murray Machine working the area? I don’t think so. The BBS model therefore consistently expects this to be a Labour hold. The SNP may also be very slightly impacted by a last minute change of candidate: not something that in and of itself causes a huge problem usually, but “party drop candidate in your seat” does filter through to voters a little bit and might give some pause.
Dumbarton (Constituency)
2021 Constituency Vote
2026 Constituency Vote (Projected)
Jackie Baillie is Holyrood’s great survivor. An MSP for this seat since 1999, it’s the only constituency Labour have never lost: Edinburgh Southern above (well, the previous version of it) went SNP in 2011 and was regained in 2016. Dumbarton however has thrice weathered national swings that by rights would have done Labour in. Instead, Baillie has clung on each time, positively devouring the Conservative and Lib Dem votes to do so.
If the SNP haven’t won this before now, I struggle to see them doing so when they’re expected to have lost a bit more than a quarter of their overall constituency vote. Even proportionally Labour are doing less badly than that, and the biggest portion of the Baillie vote that is tactical will come from the affluent Helensburgh area. I expect those voters to stick with Labour rather than switch to Reform, even if other 2021 Labour voters in Dumbarton itself and the Vale of Leven do make the switch.
East Lothian Coast and Lammermuirs (Constituency)
Notional 2021 Constituency Vote
2026 Constituency Vote (Projected)
One of just three constituencies Labour won in 2016, East Lothian (as it was then) fell to the SNP in 2021. It did so only very marginally however, though a small boundary change has removed a strongly Labour area which has further padded the SNP’s slender majority. This is nonetheless the easiest constituency for them to regain, and the model currently has them with a very small advantage that means it lies within tossup territory. If Labour don’t regain this though, they have absolutely no hope of making other gains necessary to have even a vaguely positive outcome.
Edinburgh Central (Constituency)
Notional 2021 Constituency Vote
Note: the Green vote is based solely on where they stood in 2021 and does not reflect a “true” share for this constituency.
2026 Constituency Vote (Projected)
If you’ve already read the Greens’ party profile, you’ll perhaps be familiar with this seat. For Labour it needs less explanation. Incidentally, I’ve got a little gripe about constituency naming patterns. East Lothian gets a very small change, and they stuck “and Lammermuirs” onto the end. Edinburgh Central is redrawn to such a degree it’s barely the same seat and it completely flips the 2021 result from “SNP over Conservative” to “SNP over Labour near marginal”, and it gets to keep the same name, making it easy for, ahem, dishonest incumbent MSPs to peddle nonsense about who their opponent is.
Anyway, on paper this is the second easiest constituency for Labour to gain from the SNP: given the 10.5% difference between the parties, that speaks to how far behind Labour were in 2021. The spanner in the works is that the Greens are making a very serious play for what is their best chance at a constituency MSP, in the best circumstances possible. On a good day for Labour, that splits the combined SNP and Green vote in such a way they come up the middle. Every other day though they get beaten by one of the two, and given the Greens are on the up and SNP on the down, perhaps it’ll be the Greens.
Hamilton, Larkhall and Stonehouse (Constituency)
2021 Constituency Vote
2025 By-Election Vote
2026 Constituency Vote (Projected)
These five seats have just run up the table in order of winnability, and Hamilton, Larkhall and Stonehouse is next on the list as the third most easy to gain. At 12.6% behind in 2021, that’s quite a margin, but one that Labour were able to close in last year’s by-election. The problem, as I said at the time and have kept saying since, is they only did so by 2%.
If you beat an incumbent government party, whose voter base is known to be lower turnout than yours, at a by-election with even worse turnout, when willingness to do a protest vote is at its highest, in your third most winnable seat, by 2%? That’s not a good sign! Certainly it’s far preferable to losing, but that’s too close in that context!
There’s a very real risk Labour lose this again, and if we’re being honest, they may not be helped by the fact their new MSP has been invisible since his election. I’m an avid politics watcher, obviously, and I’ve seen hide nor hair of Davy Russell since he won. Compared that to Michael Shanks who won the Rutherglen and Hamilton West by-election two years previously. They sent him all the way down to London and you could still barely move for a few months afterwards without him popping up.Â
What to Watch for in 2026
Note: Obviously, your personal perception of a good or bad result will depend on how much you like a given party. For the purposes of this piece, good and bad relate to how an impartial observer might view the result, taking into account other elections and the general situation facing that party. They are not a commentary on whether such results would be good or bad for the country.
Good Result
Not so very long ago the only possible good result for Labour was “gaining seats”. They are faring so badly in current polling though that feels an unfair expectation, and instead I’d say what passes for a good result is if they don’t lose seats at all. If they can break their trend of losing seats at every election, they can at least salvage something from this election. Obviously it’ll be better for them to make gains, even a single seat, but I’m nothing if not even handed.
Bad Result
Losing seats. Losing any seats at all. Every other party has a margin between what I’ve said is a good or bad result for them; even the Lib Dems have a single seat in between the two points where it’d feel like a relatively neutral result versus expectations. Labour does not have that, in my view. If they go to even 21 MSPs, that’s a disaster. There is no positive spin you can put on that.
You can put positive spin on no change, in the same way Sarwar put positive spin on not actually measurably doing better than Richard Leonard was polling in January 2021. You can go “ah but polls had us losing seats but we held on!” in the way Sarwar was giving it “well we’d have lost lots more seats than I did under Leonard!” But this year? “Oh well some polls had us on below 20 MSPs, but we kept losses to a minimum” rings hollow when you’ve spent the whole time since 2021 saying you’re going to be the next Scottish Government.
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)
