Note: Due to the detailed nature of the topic, this piece is significantly longer than a usual BBS article!
Boundaries Scotland have recently published the provisional proposals (link to the primary page about the review) from the second review of Scottish Parliament boundaries (link to the primary page about the review). Formerly known as the Local Government Boundary Commission for Scotland, Boundaries Scotland were renamed at the same time as they were given responsibility for Scottish Parliament boundaries, which had previously lain with the related Boundary Commission for Scotland.
Boundaries Scotland have reviewed the 70 “mainland” constituencies in the Scottish Parliament and suggested new boundaries for most of them. The three Islands seats (Na h-Eileanan an Iar, Orkney, and Shetland) are all “protected” due to their unique geographic circumstances and so are not part of the review. Although the law requires constituencies to be grouped into eight electoral regions, at this stage Boundaries Scotland have not formally done so, and will consult on the regional composition next year.
Why is it Happening?
The Scotland Act requires Boundaries Scotland to conduct a review of Scottish Parliament boundaries every 8 to 12 years. The current constituencies were first used in 2011, so legally they have to be reviewed before the next regular election in 2026, as that’s beyond the 12 year timescale. There is no option for Boundaries Scotland not to conduct a review now, and it has not been ordered or timed specifically by either the UK or Scottish Governments.
How is it Happening?
As I have written before, boundary reviews are a natural and normal part of using First Past the Post (FPTP), which is how we elected 73 of the 129 seats in the Scottish Parliament. For any voting system to be seen as fair, there needs to be a roughly equal ratio of elected members to voters in every area taking part in the same election. Whilst under PR it would be possible to have one area that is twice the size of another by giving it twice as many elected representatives, under FPTP each area has only one representative, so they should all be almost identically sized.
Excluding the three protected islands seats, the total electorate from September 2022 is 4,193,133. The commission has to use this figure for their work. With 70 constituencies on the mainland, on average each seat will have 59,902 voters within it. The rules require the commission to design constituencies as close to the average as possible, however, unlike the UK Parliament review there is no strict limitation of 5% variation from the average, allowing greater but not unlimited flexibility.
These provisional proposals are out for consultation for a month, running until the 17th of June. A second consultation on revised proposals might happen later this year or in early 2024. The proposals for regions will be consulted on in 2024, though this will be limited to fitting the final proposed constituencies within eight regions.
Boundaries Scotland do aim to keep as many constituencies to within council area boundaries as they can, but it’s not possible to do so in every case. Although FPTP requires this process to create 70 roughly equally sized chunks, Scotland does not naturally split into 70 (or any number) of equal chunks, as that simply isn’t how human population distribution and local connections work. It is therefore inevitable some constituencies will look “weird”.
Scotland and the UK do not suffer from “gerrymandering”, the drawing of electoral boundaries to fit a particular political party, the way the United States does. Boundaries Scotland, just like the Boundary Commission for Scotland, is a non-partisan and impartial body. It does not consider political advantage or disadvantage in its work. Full meeting minutes and papers available on their website can give you a fuller breakdown of their though processes.
Can We Do Things Differently?
Right now, in the moment, no. So long as the Scottish Parliament’s electoral system is the Additional Member System (AMS), which includes an FPTP element, we have to have regular boundary reviews. However, boundary reviews are a disruptive process which highlight the intrinsic contradictions of FPTP. We’re often told in the UK that we have to have FPTP in order to give voice to “local” areas, and yet we chop and change the lines for what we consider local to fit electoral needs. If two towns together form a constituency one election, then are split at the next, how can you credibly claim that they were or are being given proper local representation?
Many other countries in Europe lack even the very concept of boundary changes, because they don’t use FPTP or other systems electing very small numbers of people per area, such as the Single Transferrable Vote (STV) which we use for council elections. Instead, most countries in Europe use forms of Party List Proportional Representation (PLPR). Whilst this is less local, it allows for greater consistency between elections and between different layers of governance.
It also helps in most of Europe that they have genuine local government to deal with local issues. In Scotland (and the wider UK) we are in the bizarre circumstance of swearing blind that e.g. East Kilbride needs a specific single representative in parliament, but no, it certainly doesn’t need its own burgh council and can instead be part of the sprawlingly un-local South Lanarkshire. I’d say we have things completely back to front here, but that’s even more of a tangent… that I’ll be revisiting later this year or early next with a new version of my New Municipalism project.
Most countries using PLPR will use their provinces, counties, regions, or whatever they may call major subnational divisions as voting areas. That means each area has some degree of recognition and basis in either historic or current layers of government. Instead of changing the boundaries between elections, they will change the number of seats each area has. If the population has increased in a province that currently elects 11 MPs, it might be increased to 12. Meanwhile, if there’s a decrease for a province with 7 MPs, it might drop to 6. This achieves roughly the same goal as boundary changes, without the disruption.
I’m using a slightly different approach to describing geographic changes than Boundaries Scotland did, by outlining whether changes are minor, major, or effectively result in a new constituency entirely.
- Unchanged – Self-explanatory!
- Minor – Only a minor change to an existing seat, which may include a change of name.
- Major – Major change to an existing seat, which nonetheless remains recognisable in some form.
- New – A constituency with big enough changes it’s worth viewing as a new seat entirely.
Although it’s not in any way the point of the exercise, it’s an inevitable fact that if you change electoral boundaries, that has an impact on the political balance of the changed seats. However, the fact that the Scottish Parliament is partly proportional means that boundary changes are both likely to be less consequential than under pure FPTP, and less direct.
When AMS is working to perfection, changing a constituency between parties without any difference in regional vote just triggers an equal but opposite change in list seats. So for example, the Conservatives winning an additional constituency in the South region doesn’t actually give them an additional seat, as they receive one fewer list seat, and the SNP get one more.
When AMS isn’t working to perfection, because a party has exceeded its proportional share via constituencies, it may not be the party losing another constituency which loses a seat. As another example, the SNP already had one seat more than their proportional share in the North East region in 2021, costing the Conservatives a list seat. If they’d also won the one constituency the Conservatives won, then the compensating list seat for the Conservatives would come from Labour, who had won the 7th and final list seat otherwise.
It’s also worth bearing in mind the SNP’s dominance at the 2021 election. If you take a handful of SNP-held constituencies and redraw the boundaries between them… you still have a handful of SNP-held constituencies. Winning margins may fluctuate, but only in a very small number of cases is that likely to take comfortably SNP seats into marginal territory. At the same time, any opposition-held constituencies that are redrawn to include parts of SNP-held seats will almost certainly find the SNP gaining ground, given the concentration of votes within the current boundaries.
With all of that in mind, I’ve done some rough estimates of impacts on the most interesting constituencies, which work roughly on the following basis:
- Using the unparalleled detail available in the polling district level results for the 2022 local elections, reconstruct an existing constituency in 2022 terms.
- Work out the distribution of votes for each party across the whole constituency.
- Assume 2021 is likely to have had a relatively similar distribution.
- Move 2021 votes in the same proportion, so if e.g. 10% of a party’s 2022 vote in a given constituency came from the portion being moved into a new constituency, move 10% of their 2021 vote accordingly.
This is not a perfect process, especially given the fact that no party stood in every ward in 2022 and so there are some gaps, but it’s better than guessing and it’ll do for the moment! I am not doing this for absolutely every seat at the moment just because it would be an absolutely massive task, but I will do it when we have final proposals published.
The maps below show a simple side by side comparison of current and existing constituencies. Although Boundaries Scotland have not grouped constituencies into regions at this stage, I have given two scenarios for my best guess at regional configurations. Scenario 1 is the one I consider most likely, but Scenario 2 was pointed out by a Twitter follower and whilst I think it’s less likely it is the only alternative arrangement I think would be at all likely.
That overall picture out of the way, let’s delve right into the detail of the proposals! I’ll take each (likely) region in turn, with the grouping based on the first scenario but commentary covering both. I am not providing the map for each constituency because you can get those from Boundaries Scotland here.
There are some pretty significant changes within the Central region, with only the Falkirk seats unchanged entirely. This is also one of three regions where there is a degree of uncertainty about what the exact makeup would be, with the option I consider most likely involving losing Kilsyth to West whilst gaining most of Clydesdale from South. The alternative gains most of East Dunbartonshire from West whilst losing Larkhall to South.
Airdrie, Newmains and Shotts (Minor)
A relatively simple expansion of the current Aidrie and Shotts constituency, drawing in the village of Newmains plus the Wishaw suburbs of Cambusnethan and Coltness from the current Motherwell and Wishaw.
Bellshill and Coatbridge (New)
The name basically sums this one up very nicely – it takes Bellshill from the current Uddingston and Bellshill and Coatbridge from the current Coatbridge and Chryston. That means no constituencies cross the boundary between North and South Lanarkshire anymore.
Cumbernauld and Chryston (Major)
Another simple enough change, it’s Cumbernauld from the current Cumbernauld and Kilsyth plus the Chryston, Stepps, Muirhead, Gartcosh, Glenboig and so on portion of the current Coatbridge and Chryston. This isn’t a huge change in number of voters as Cumbernauld is the bulk of it, but the separation of Cumbernauld and Kilsyth is a major change.
East Kilbride (Minor)
Gains a very small area to the southwest, where some new housebuilding has been going on, from the current Clydesdale constituency.
Falkirk East (Unchanged)
Falkirk West (Unchanged)
Hamilton and Uddingston (Major)
As the name implies, takes the Hamilton from the current Hamilton, Larkhall and Stonehouse and the Uddingston from the current Bellshill and Uddingston. Hamilton is the majority of the constituency either way.
Larkhall and Clydesdale (New)
This is the bulk of the population but not the area of the current Clydesdale constituency, including Carluke, Strathaven, Lesmahagow and Douglas, amongst other bits, as well as the Larkhall and Stonehouse bits of the current Hamilton, Larkhall and Stonehouse.
This is one of the three constituencies that could have be placed in an alternate region for scenario 2, which would be South. However, it does not make a difference to estimated 2021 seat shares either way.
Despite the name change, this does still include most of Wishaw, despite the earlier mentioned losses to the redrawn Airdrie, Newmains and Shotts seat. It also gains some of Motherwell’s northern suburbs from the current Uddingston and Bellshill, around Carfin, Newarthill, New Stevenston and Holytown.
Regional Impact - Scenario 1
The overall boundaries of the region would have to change to account for new constituency boundaries. The core scenario I’m working to includes the constituencies outlined above, on the basis that Larkhall has a larger population than Kilsyth, and thus keeping the constituency with Larkhall in it, and absorbing a large portion of the current Clydesdale constituency from the South region, results in the least change to current regional boundaries.
In 2021, the Greens scraped the last seat in the region here by just 107 votes over a fourth Labour seat, which would be very susceptible to possible changes. In this scenario, the loss of Kilsyth benefits the Greens somewhat as it’s a very strongly Labour area (one of their best wards in both 2017 and 2022) and so costs Labour a bit more dearly, and the Clydesdale component doesn’t make up for it, and I therefore estimate the Greens keep that seat.
Regional Impact - Scenario 2
The second scenario would see Larkhall and Clydesdale in South, and so instead Central gets Kirkintilloch and Kilsyth. You might think for a second given what I said about Kilsyth earlier that keeping it in Central helps Labour, but no, not when it’s now bringing most of East Dunbartonshire with it. I estimate the Green vote in that side of the proposed constituency to be 8%, significantly above the 6% they won across Central. With the Conservatives also weaker without that Clydesdale proportion, this actually sees the Greens pick up the 6th regional seat, and it’s the Conservatives who pip Labour to the 7th.
Despite the name, at present the Glasgow region includes a constituency entirely within South Lanarkshire, albeit the bulk of it had been part of the city between 1973 and 1994. The proposed changes are so significant however that not only is the internal division of seats in the city massively different, but one of them also reaches across current regional lines to draw in part of East Renfrewshire. Although this may be a shock for some, having had a quick look at the relevant minutes and discussion papers, this is probably the least disruptive option.
The only way to both give Glasgow 8 constituencies and to prevent a much more significant break between the Renfrewshire-Ayrshire boundary is to have a Glasgow-East Renfrewshire seat. Avoiding crossing both the Glasgow and the Renfrewshire-Ayrshire boundary requires Glasgow to have only 7 constituencies. If Glasgow only has 7 constituencies, it will need 1 more after Rutherglen to get up to a full regional complement of 9 seats, which means a much more dramatic crossing of the city boundary at regional level than already exists.
Glasgow Anniesland (Minor)
Gains Thornwood from the current Glasgow Kelvin, loses Claythorn to the redrawn Kelvin and Maryhill.
Glasgow Cardonald and Pollok (Major)
Notable as the largest portion of this is drawn from the current Glasgow Pollok seat held by the current First Minister, it also draws in Pollokshields, Strathbungo and part of Shawlands from the current Glasgow Southside.
Glasgow Central and Govan (New)
None of the current Glasgow constituencies cross the river, but this new seat consisting of parts of four existing constituencies does so. From the current Glasgow Kelvin it draws the Anderston, Yorkhill and City Centre ward, from Glasgow Shettleston it pulls in Calton and Bridgeton, from Glasgow Southside it’s got Oatlands, the Gorbals, Kinning Park and Ibrox, and finally from Glasgow Pollok it takes Govan. One of the Kelvin successors, it doesn’t have the same concentration of Green vote as that does.
Glasgow Kelvin and Maryhill (Major)
From the current Glasgow Kelvin this only retains Hillhead, Hyndland and Dowanhill. It gains Claythorn from Glasgow Anniesland, and then areas like Maryhill, Lambhill, Ruchill (lots of -hills here, how positively Roman) and the western part of Possil from Glasgow Maryhill and Springburn. The Greens will be absolutely scunnered with this one, as they’d tried to win the current Glasgow Kelvin at the past two Holyrood elections, coming second both times.
The existing constituency consists almost entirely of strong Green areas, but being split four ways dilutes that significantly. Although the core here contains the best Green ward in the country at Hillhead, Kelvindale, Maryhill and Ruchill just cannot compare areas like Thornwood, Finnieston and Garnethill. I haven’t done formal estimates because for this I’d need to further try and estimate Green shares in the bits they didn’t stand in, but this one is unlikely to be Green-winnable in this state.
Glasgow Priesthill and Giffnock (New) - Key Seat
If crossing the Clyde wasn’t a big enough difference from current Glasgow constituency boundaries, how about crossing the City Council boundary itself? This draws the Priesthill and Nitshill parts of the current Glasgow Pollok together with the Newlands and Carnwadric parts of the current Glasgow Cathcart, and crosses into the current Eastwood constituency (and West region) to take in Giffnock, Thornliebank, Clarkston, Busby and more. Although they’d never be able to translate it to a win, the inclusion of the latter would probably make this the strongest “Glasgow” constituency for the Conservatives.
Glasgow Shettleston and Baillieston (Major)
The bulk of the current Glasgow Shettleston bar the Calton and Bridgeton bits, it also draws in areas like Riddrie, Carntyne, Cranhill and Springboig from the current Glasgow Provan. Although the name may suggest otherwise, the current constituency does actually include Baillieston – I think it’s just been added to differentiate based on the boundaries changing.
Glasgow Southside and Cathcart (Major)
Don’t let the name fool you here – the contribution from the current Glasgow Southside is very limited, just Govanhill and Polmadie. It’s Glasgow Cathcart that contributes most to this, through Battlefield, Langside, Cathcart and Castlemilk, even with Newlands and so on going to the proposed Glasgow Priesthill and Giffnock seat.
Although the Linn component is not exactly Green favourable, especially around Castlemilk, if they were looking for a “new Kelvin” in the city to try and win, this might not actually be a bad shout. In 2022 they did extremely well in the Langside component, pushing Labour into third place, and the west of Govanhill was very good for them too.
Glasgow Springburn and Provan (Major)
Most of this comes from the current Glasgow Provan, namely Easterhouse, Garthamlock, Robroyston, Provanmill and Dennistoun. From Glasgow Maryhill and Springburn, this has the east of Possil, Milton, and Springburn itself.
This really is basically just the current Rutherglen seat, bar one little housing estate in Blantyre.
In addition to the blowing Kelvin to smithereens, the Greens also won’t be best pleased with Giffnock and Clarkston being added to the region. That area has a good Green vote by the standards of West, but a weaker Green vote by Glasgow standards. Throw in the strong favourability for Conservatives, and it makes it a bit harder for the party to pick up a second MSP in both regions. As that change is also unfavourable for Labour, rather than coming about 915 votes off beating the second Conservative, the estimate here is the Greens end up 1071 votes short of beating the fourth Labour candidate.
Highlands and Islands
As the three Islands constituencies are not included in boundary changes, combined with only some very minor changes, Highlands and Islands would be the second least altered region overall.
Argyll and Bute (Minor)
Argyll and Bute has the lowest electorate of any existing unprotected constituency, so it’s very hard to justify not expanding it. It does so by taking Lomond North from the current Dumbarton (and thus West region), though it’s worth noting that in local council terms, that’s all part of the Argyll and Bute area anyway, even if it was historically (and correctly) Dunbartonshire. Folk in Lomond North may well end up strenuously protesting their separation from Helensburgh here though.
Caithness, Sutherland and Ross (Unchanged)
Still the geographically biggest constituency in the country.
Inverness and Nairn (Minor)
Changes here look a lot more substantial than they are, given the nature of Highland geography. In population terms this really isn’t a big change, just losing Nethy Bridge and Boat of Garten.
Na h-Eileanan an Iar (Unchanged)
Orkney Islands (Unchanged)
Shetland Islands (Unchanged)
Skye, Lochaber and Badenoch (Minor)
Just picks up the bits Inverness and Nairn lost, amounting mostly just to Nethy Bridge and Boat of Garten.
As only a small area is added to the region and the final list seat wasn’t at all marginal, there are no significant impacts.
If you thought Glasgow was messy, strap in. Population growth in the Lothians has been very strong, and that is what leads to some very dramatic changes here. The current boundaries are already in the situation of having the Lothians spilling out of the Lothian region – with 9 constituencies in the region, plus one wholly and one half Lothians seat in South. These proposals would inflate to 12 seats entirely within the Lothians, and by the only reasonable possible makeup, 10 within the Lothian region.
Bathgate and Almond Valley (New)
West Lothian was previously split North-South between Linlithgow and Almond Valley, but no more. This new seat takes Bathgate, Armadale, Whitburn and Blackburn from the current Linlithgow, whilst Seafield, West Calder, Fauldhouse and the general Breich Valley area come from Almond Valley.
Edinburgh Central (Major) - Key Seat
The current Edinburgh Central caused one of the biggest shocks of 2016 when then-leader of the Scottish Conservatives Ruth Davidson gained it from the SNP. It was a hard-fought battleground in 2021, with the SNP eventually prevailing quite comfortably. This… is not really that seat. From the current Edinburgh Central this does indeed contain the city centre, both Old and New Towns, as well as Tollcross, Canongate, and what the map I’m referring to simply calls “Southside”.
However, it adds Newington from the current Edinburgh Southern and Broughton from the current Edinburgh Northern and Leith. The removal of strong Conservative areas and shift onto a bit of Labour’s current Edinburgh Southern patch makes it an SNP-Labour marginal in 2021 terms, with the Conservatives a distant third.
Edinburgh Eastern (Major)
This loses Portobello and Brunstane, which we’ll get onto later, whilst gaining Liberton.
Edinburgh Forth and Linlithgow (New) - Key Seat
People already don’t like this one, but I’m afraid that you’re just going to have to suck some form of it up. See the whole “Lothians are growing rapidly” thing? Yeah, part of that is that West Lothian is now due about two and a half constituencies at Holyrood. It just can’t be contained in two anymore. Like it or not, tying part of West to Queensferry (historically in West Lothian) is probably the most sensible play.
From the current Edinburgh Western this takes basically the whole Almond ward, including South Queensferry, Kirkliston, Carmond and Muirhouse. Meanwhile, the current Linlithgow provides Linlithgow itself plus Broxburn, Uphall and Winchburgh. Almond wasn’t just the Lib Dems’ strongest ward in Scotland last year, but the strongest ward for any party. Combined with solid support in Linlithgow, folk in the Broxburn conurbation might be completely stunned to find themselves in an SNP-Lib Dem marginal.
Edinburgh Northern and Leith (Minor)
Loses Broughton to the newly redrawn Central and the core Inverleith to the proposed Edinburgh Western, but otherwise one of the least-scathed Edinburgh seats.
Edinburgh Pentlands (Minor)
Gains Slateford from the current Edinburgh Southern and western Gorgie from the current Edinburgh Central, plus the very small contribution of Ratho Station from the current Edinburgh Western. Loses Oxgangs and Fairmilehead to redrawn Edinburgh Southern.
Edinburgh Southern (Major) - Key Seat
A significant reconfiguration here, taking what was an east-west oriented seat and making it north-south. Keeps Morningside, Merchiston and Craiglockhart. Gains Dalry, eastern Gorgie and a wee bit of Fountainbridge from the current Edinburgh Central, though it loses Newington to the redrawn version of that. Picks up Oxgangs and Fairmilehead from Pentlands, but has then lost Liberton to redrawn Eastern.
One of Labour’s only two remaining constituencies, this could have been very vulnerable to a major redraw. In actuality, it mostly moves onto territory that’s still pretty good for Labour. That means my estimate here is for Labour to have won it in 2021, albeit on a very small margin – and a margin that’d probably have been even smaller when removing the methodological artefact of a small Green vote coming from the current Central portion.
Edinburgh Western (Major) - Key Seat
A huge change to Edinburgh Western, which only retains the Corstorphine, Clermiston, Blackhall and Drylaw portions of the existing seat. It picks up the core Inverleith from Edinburgh Northern and Leith, but most of the addition comes from the current Edinburgh Central, including Murrayfield and Comely Bank. My 2021 estimate is therefore for a very narrow SNP win over the Lib Dems, though again if we account for the fact the Green vote is only coming from the Edinburgh Central component and wouldn’t have been there in 2021 it might have been a bit wider.
However, Alex Cole-Hamilton and the Lib Dems don’t really have anything to fear from this one, not least because he’d have secured a list seat instead. Beyond the fact that a real 2021 election on this boundary would have seen them ramming some tactical vote messaging home to Conservative voters, the Lib Dems were completely dominant in the wards making this up in the 2022 elections. I’d be absolutely stunned if this wasn’t an easy Lib Dem hold in 2026 – however, that’s not the basis we work boundary change impacts on!
In short, given the Lib Dems could win both of the constituencies coming out of the current Edinburgh Western, regardless of what they might say about it publicly, internally they’ll be laughing.
I’m considering this a major change from the current Almond Valley constituency, which truncates right down to just Livingston itself. That’s no surprise given Livingston is where a lot of the new house building and population growth has been.
Midlothian North and Musselburgh (Major)
Another one people really, really won’t like. The core of this remains the Musselburgh, Wallyford, Dalkeith and Danderhall areas from the existing constituency. However, the Midlothian component is heavily reduced, as the likes of Loanhead, Bonnyrigg and Mayfield pass out of the constituency (and the region) into a Midlothian South seat. Where does the difference come from? On the one hand, adding Prestonpans from the current East Lothian. On the other, and perhaps controversially, by pulling Portobello and surrounds like Brunstane and Joppa out of the current Edinburgh Eastern.
The fact the SNP end up estimated to win two more constituencies in this region on changed boundaries has a significant impact. Since it creates and overhang and pushes the Lib Dems onto a list seat, whoever would otherwise be up for seat 7 is in the firing line. In 2021 that was the Greens, and whilst their overall share in the region ticks marginally upwards due to the removal of comparatively weaker Midlothian areas, my estimate comes to 119 votes short of beating the third Labour candidate to the last seat. Labour may not want to protest too much about East Lothian losing Prestonpans, as that’s what makes the difference for them here.
Mid and Fife
Compared to some of the other regions, Mid Scotland and Fife is relatively quiet. There are more significant changes in-region than there are in Highlands and Islands so even though there are no changes to the overall regional borders, I’d put this as the third-least changed.
Clackmannanshire and Dunblane (Minor)
There’s a clear attempt here to beef up the numbers without things getting too awkward, and it does that by adding the area around Doune and Thornhill from the current Stirling. As a noted fan of the Wee County, I will however say there’s something about the basic geometry of this constituency I find unsettling, which is a very weird reaction to have to a constituency map.
Loses Cardenden and vicinity to the redrawn Mid Fife and Glenrothes, whilst gaining Burntisland from the current Kirkcaldy and Charlestown and Limekilns from the current Dunfermline.
In case you skipped the previous line, it’s lost Charlestown and Limekilns to the redrawn Cowdenbeath.
Fife North East (Unchanged)
The order of the name has shifted though.
Loses Burntisland to the redrawn Cowdenbeath.
Mid Fife and Glenrothes (Minor)
Gains Cardenden from the current Cowdenbeath.
Perthshire North (Minor)
Gains Almondbank from the southern seat.
Perthshire South and Kinross-shire (Minor)
Loses Almondbank to the northern seat.
Already not the full council area in one seat, it’s trimmed down by the loss of Doune and Thornhill to the redrawn Clackmannanshire and Dunblane.
Since none of the constituencies would be any different, the SNP’s overhang isn’t rectified, and since the region doesn’t change either, it’s all the same as the actual 2021 vote.
The North East wins the award for “least changes” as there is one very, very small tweak between two constituencies. Everything else is exactly the same.
Aberdeen Central (Unchanged)
Aberdeen Donside (Unchanged)
Aberdeen South and North Kincardine (Unchanged)
Aberdeenshire East (Minor)
Loses Strichen, New Pitsligo, New Leeds and Fetterangus to Banff and Buchan Coast. It also loses the northern portion of Longside, a village of just 1,030 people. This is weird and fixing it wouldn’t cause any detriment to the electorate numbers so for the sake of tidiness, you’d be doing me a favour if you told Boundaries Scotland to fix it. This was a marginal in 2021 but the changes are really tiny so I haven’t charted them. They move the dial very slightly in the SNP’s favour, going from a 4.6% to 5.2% margin.
Aberdeenshire West (Unchanged)
Angus North and Mearns (Unchanged)
Angus South (Unchanged)
Banff and Buchan Coast (Minor) - Key Seat
In addition to losing the “-shire” from Banff, perhaps to emphasise the slight change in boundaries, but this picks up the bits from Aberdeenshire East – Strichen, New Pitsligo, New Leeds, Fetterangus. The current Banffshire and Buchan Coast was the second most marginal seat in the country at 2.3%, so even small changes could make a difference. However, these are way too small, so it narrows to 1.9% in 2021 terms but still SNP.
Dundee City East (Unchanged)
Dundee City West (Unchanged)
No change to constituency winners, no change to region boundaries, no change to MSPs.
Most of the constituencies in South go unchanged, and even if they had, given it starts with the most non-SNP constituencies, changing constituency winners alone wouldn’t be likely to change the total number of seats, as the lists would be more capable of delivering proportionality. However, changes to the regional boundary are sufficient to deliver one clear change versus current boundaries.
This is also one of three regions where there is a degree of uncertainty about what the exact makeup would be, with the option I consider most likely involving losing most of the population of Clydesdale to Central. The alternative gains Larkhall from Central whilst losing Kilmarnock to West.
Remains the most marginal seat in the country at 0.4% SNP over Conservative.
Carrick, Cumnock and Doon Valley (Unchanged)
Clyde Valley and Tweeddale (New) - Key Seat
This one takes the bulk of the area of the current Clydesdale, including Lanark, Forth, Biggar, and Abington, plus the bulk of the population of the current Midlothian South, Tweeddale and Lauderdale constituency. Whilst crossing the South Lanarkshire-Scottish Borders, err, border might seem like a bold move, remember that there’s a UK Parliament constituency similarly covering parts of both Clydesdale and Tweeddale. As a Central Belter through and through it does often boggle my mind that Lanarkshire borders both Glasgow and the Borders, but it does!
In most cases, putting parts of comfortably SNP-held seats together naturally gives you a comfortably SNP-held seat. Not so here, because these components are the most Conservative-friendly parts of their current constituencies. That creates a new, very competitive SNP-Conservative marginal. That margin would likely be even narrower when considering Green votes coming from the Clydesdale side – it was very much the Borders branch of the Greens who wanted to contest the current MSTL seat, so they’d certainly have been on the ballot in this one instead.
East Lothian (Minor) - Key Seat
Loses Prestonpans to the redrawn Midlothian North and Musselburgh constituency, which also takes that area out of the South region. On current boundaries, this was the third most marginal seat in the country, with the SNP only 2.6% ahead of Labour. Prestonpans is much more strongly Labour however, so my estimate here is that removing it boosts the SNP’s margin to 4% in 2021 terms.
Ettrick, Roxburgh and Berwickshire (Unchanged)
Galloway and West Dumfries(Unchanged)
Kilmarnock and Irvine Valley (Minor)
Loses Stewarton, Kilmaurs and Dunlop to Cunninghame South, and therefore also to the West region.
This is one of the three constituencies that could have be placed in an alternate region for scenario 2, which would be West. However, it does not make a difference to estimated 2021 seat shares either way.
Midlothian South (New)
This is about five-sixths of Midlothian, as all of Midlothian’s wards are three-seats for local elections and only one of them isn’t in this constituency. Only Dalkeith ward isn’t within the constituency, and therefore not within the South region.
Regional Impact - Scenario 1
As with Central, the boundaries of the region overall would have to change to accommodate constituency changes here. My instinct is that the version with Kilmarnock in it would be most likely, as it results in the least change relative to current regional boundaries, as Kilmarnock has a larger population than the Clydesdale component it would keep otherwise if Central didn’t get Larkhall and Clydesdale.
In 2021, South was the only region the Greens failed to win a seat in. They came extremely close however, at only 115 votes away from displacing the third Labour MSP. Regardless of which scenario applies here, the changes around the Lothian edges of the region are sufficient to make up that difference. That’s due to a combination of losing Prestonpans, a strong Labour but middling Green area, whilst gaining that middle strip of Midlothian. The Green share in the latter estimates to 7.8%, comfortably above the 5.2% they won in the current region.
Regional Impact - Scenario 2
In the second scenario, Kilmarnock and Irvine Valley goes to West, and Larkhall and Clydesdale is in South instead of Central. This actually has the effect of further boosting the Greens, as Clydesdale is a fair bit weaker for the SNP than Kilmarnock is. In fact, the Greens go from no MSP in the region at all to end up with the 6th seat in this scenario, with the SNP taking the 7th.
West ends up being the counterbalance to Lothian. Whereas the population across Lothian has been increasing rapidly, not so much across Renfrewshire in particular. Not only do most constituencies here have changes, but there’s no configuration to my mind that doesn’t lead to a 9 constituency region, down from 10.
This is also one of three regions where there is a degree of uncertainty about what the exact makeup would be, with the option I consider most likely involving gaining Kilsyth from Central. The alternative would be losing most of East Dunbartonshire to Central, whilst gaining Kilmarnock from South.
Bearsden, Milngavie and Clydebank (Major)
This is effectively a redrawn Clydebank and Milngavie constituency. Although the name might have made you think otherwise, the current seat actually already has half of Bearsden in it, and this version gains the missing Bearsden South ward. Just to confuse things though it also loses a large portion of the Clydebank Waterfront ward to a redrawn Dumbarton, stretching from just by the town centre out to Old Kilpatrick.
Cunninghame North (Unchanged)
Cunninghame South (Minor)
Gains the Stewarton, Kilmaurs and Dunlop area from the current Kilmarnock and Irvine Valley, and consequently from the South region.
Dumbarton (Minor) - Key Seat
Loses the Lomond North area to the redrawn Argyll and Bute and thus to Highlands and Islands region, whilst gaining most of the Clydebank Waterfront ward from Old Kilpatrick through almost to the town centre. Labour MSP Jackie Baillie is a consummate survivor, and three times so far has held fast against national swings that if replicated locally would have seen her relegated to the list – or, in their 2011 disaster, ejected entirely. In 2021, she managed to grow her majority to a still slender 3.9%.
Given Lomond North will be a rich source of tactical votes for Baillie whereas Clydebank Waterfront is one of the most SNP-leaning wards in the country, I had wondered if even that minor change would be enough to narrowly tip this the SNP’s way. Answer? Not quite, with Labour maintaining an estimated 1.5% lead. Jackie Baillie, it seems, cannot be beaten, either by the SNP or Boundaries Scotland!
The current name of “Greenock and Inverclyde” is a bit weird, seeing as it doesn’t cover the whole of Inverclyde. This version does, having added Kilmacolm and Quarrier’s Village.
Kirkintilloch and Kilsyth (Minor)
Though the name changes accordingly, this is a relatively mild redrawing of the current Strathkelvin and Bearsden, removing Bearsden South but adding Kilsyth.
This is one of the three constituencies that could have be placed in an alternate region for scenario 2, which would be Central. However, it does not make a difference to estimated 2021 seat shares either way.
Paisley and Renfrew (Major)
Compared to the current Paisley constituency, this loses the Paisley Northwest ward which covers areas such as Castlehead and Ferguslie Park. It instead gains Gallowhill and Renfrew. I have to admit I really don’t like this one. The current Paisley seat doesn’t have all of Paisley in it since it’s lacking Gallowhill, but that just looks a lot less horrendous than this. Gallowhill is also only about half of a three member ward, whereas Paisley Northwest is a full four member, so this has a net loss of Paisley.
I’m going to try play around with numbers for my own consultation response, but I think even swapping Renfrew North and Braehead to the proposed Renfrewshire West for Paisley Northwest would be marginally less awful. That said, I acknowledge it means a proportionally much larger split in Renfrew, which may be why Boundaries Scotland opted for this instead.
Renfrewshire South (Major) - Key Seat
This is one of the most consequential proposed changes in the country, as it’s part of the dissolution of the currently Conservative-held Eastwood seat. Newton Mearns and Eaglesham are drawn from that and added to the existing Barrhead, Neilston, Lochwinnoch and Johnstone core of the current Renfrewshire South. However, it loses Elderslie, Linwood, Brookfield and the eastern portion of Crosslee. Although it didn’t prove as bad for them after doing the estimate as I had thought on eyeballing, it is indeed the case that getting rid of Eastwood like this means the Conservatives lose their one West constituency in 2021 terms.
Jackson Carlaw may wish to keep such a prospect in mind before he does any further tweets seeming to suggest list MSPs are “rejected” by voters or in some way illegitimate, though this would nonetheless be a winnable marginal, albeit more from if the SNP collapse in just the right way for the Conservatives to win a three-way split rather than people in, err, Barrhead or Johnstone going all in for the Conservatives.
Renfrewshire West (Major)
A reformulation of the current Renfrewshire North and West, it picks up what the others have lost – Elderslie, Linwood, Brookfield and part of Crosslee from the redrawn Renfrewshire South, and Paisley Northwest from the current Paisley. On the other hand, it loses Renfrew and Gallowhill to the new Paisley and Renfrew, and Kilmacolm and Quarrier’s Village to the new Inverclyde.
Regional Impact - Scenario 1
Again, this is one where the boundaries of the region overall would have to change to accommodate constituency changes here. I think the version with Kilsyth in it would be most likely, as it results in the least change relative to current regional boundaries, whereas the Kilmarnock-inclusive version means losing a huge whack of East Dunbartonshire.
The fact the Conservatives wouldn’t have a constituency under these proposals bodes ill for them. In 2021, the SNP didn’t have an overhang in West only because both Labour and the Conservatives won a constituency. With the Conservatives down one, they effectively have to take a list seat from themselves, leaving them one worse off. That also makes this a bit of a different story from the other regions where it was the Greens who might have had reasons to be worried about boundaries shifting.
Regional Impact - Scenario 2
In the second scenario, West gets Kilmarnock and Irvine Valley goes to West, and Kirkintilloch and Kilsyth goes to Central instead. This doesn’t have any effect on notional 2021 results, but the Greens and Lib Dems would both prefer not to get this way around, as it means losing one of their best areas. Whilst the Greens might take consolation that it would pad out their marginal Central seat without really risking their West seat, for the Lib Dems it would leave them significantly further from gaining a seat here whilst pumping useless votes into hands-down their worst region.
Notional National Result
Regardless of scenario, my estimate is that these boundary changes gives the same number of seats.
Although changes overall are very mild with just two seats changing hands, it really matters where those seats are. In 2021 as it happened, the SNP had a single seat overhang in four regions – i.e. they knocked another party out of a seat they should proportionally have had in those regions. The other four they got their proportionally correct number of seats overall. Here, two of those four also end up with overhangs, and that takes the SNP from one seat short of a majority to one seat more than a majority. The ability to legislate on their own without external support would likely have created a very different feel to this term, not least for the Greens.
Of course, my estimates here are only estimates. I’m very unlikely to have gotten the notional figures exactly correct, due to this country’s unfortunately limited collation and publication of detailed election data. However, what is key for the SNP ending up with a majority here is that Conservative-held Eastwood in West disappears. It’s simply impossible that with 2021 votes as they were the Conservatives would have won the Renfrewshire South replacement seat, and that overhang alone gives the SNP their 65th seat regardless of what happens elsewhere.
If Edinburgh Western was really a Lib Dem seat on these boundaries, that only costs the SNP their 66th – whilst adding a 9th Green seat to the mix. In fact, the Edinburgh Southern estimate is marginal enough that could be their 66th instead (and bar a second Green again) if my figures are a bit off there. If the Conservatives were to win the Clyde Valley and Tweeddale seat it’s immaterial, as South is where the SNP were furthest from overhanging, so they just get another list seat and there’s no net difference. Changes elsewhere just aren’t sufficient to flip seats the SNP held in 2021.
Responding to the Consultation
You can respond to the consultation on Boundaries Scotland’s website here. Remember that the least useful, most absolutely pointless thing you can do is send unconstructive complaints or abuse to the commission. That helps nobody, least of all you or other people in the area you’re unhappy about. If you think there is something wrong, suggest a viable alternative – viable being the key word! Suggesting they remove a few hundred voters so they can add thousands more isn’t going to stack up. You have until the 17th of June to respond.
The Ballot Box Scotland response to the consultation is available here.
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.