2020 in Review – Council By-Elections

Background and Caveats

Well. 2020 has been a year, eh? Although in one respect it is ending as it began, with Brexit in the headlines (and causing headaches), our lives have been changed almost beyond recognition by the Coronavirus pandemic. Elections weren’t immune to the virus, and across the UK, all scheduled elections were put on hold – including the London Mayoral and Assembly elections, plus a whole rake of councils across England.

Here in Scotland the effects were milder as it just shuffled by-elections about, though I was disappointed the apparently annual March by-election in Clackmannanshire was postponed the day before. Had it not been for the pandemic we’d have at least equalled last year’s tally of 15 by-elections, but we instead ended up with just 10 as multiple postponements knocked some by-elections into 2021.

As ever, before we delve into the overall results from this year, it’s important to remember just 10 of Scotland’s 354 wards cannot be representative of the whole country. Comparisons are only against those same wards. Per my arbitrary ward classification scheme, we had 4 City, 3 Islands, 2 Urbanised and 1 Rural ward go to the polls this year. Bear in mind too that where popular Independent Councillors were elected in 2017, their obvious absence from by-elections limits the comparisons between the 2017 and by-election results.

Finally, the most crucial STV caveat is to remember that the language of “gain/loss/hold” isn’t as clear-cut as it is under FPTP, where it indicates both change in seat and change in lead party. It’s entirely possible for the “defending” party, the party that had a first preference lead, and the party that would have won a single seat election in 2017 to all be different!

Seat Shifts

We’ll delve straight into that sticky side of STV by looking at how seats have shifted. For each by-election in the table below, the name in the first column links to the preview post, and the winning party in the last column to the results post, so you can get the full detail. 

[supsystic-tables id=12]

So, if we first look at things in terms of change in councillors (Terminology differs for Independents as they aren’t directly comparable):

SNP - Vacated 5, held 4, lost 1 (Net -1)
Conservative - Vacated 2, held 2 (Net nc)
Liberal Democrat - Vacated 0, won 1 (Net +1)
Independent - Vacated 3, won 3 (Net nc)

Comparing instead with the notional 2017 winner for a single seat:

SNP - Defending 3, gained 1, held 3 (Net +1)
Conservative - Defending 3, lost 1, held 2 (Net -1)
Liberal Democrat - Defending 1, held 1 (Net nc)
Independent - Expected 3, won 3 (Net nc)

The SNP’s net figure effectively flips between those two measures, as whilst five vacancies were caused by SNP councillors, two of those were in wards the SNP would not have won a single-seat election in 2017. As they won one of those two and lost the other, they are down one seat in strict councillor terms, but up one versus their notional 2017 performance.

For the Conservatives, though they held both of the seats they caused a direct vacancy in, they were also the expected 2017 winner in the Ellon and District ward, which they narrowly lost to the SNP in the by-election. That’s why they are net-zero on the first measure but down a seat on the second.

Finally, though the Lib Dems have an additional Perth & Kinross councillor thanks to Perth City South, they would have been expected to have won a single seat in that ward in 2017 – at the full election, anyway. A by-election later that year went Conservative, so as an added layer of complexity it could be thought of as a Lib Dem gain versus that one. STV is wild, eh?

First Preferences

Image Version

We can also look at first preferences across the whole, though with another one of those huge, great big “NOT REPRESENTATIVE OF SCOTLAND AS A WHOLE” caveats up in neon lights:

SNP - 12106 (35.8%, +4.9)
Conservative - 9730 (28.8%, +2.8)
Liberal Democrat - 3845 (11.4%, +1.8)
Independent - 2981 (8.8%, -8.7)
Labour - 2745 (8.1%, -4.3)
Green - 2271 (6.7%, +3.2)
Other - 121 (0.4%, +0.4)

Though there is no sugar coating Labour’s performance either in these by-elections or current polling, the fact they come in behind the Lib Dems in this year’s by-elections neatly makes that “not representative” point. We’d have seen a smaller gap, and quite possibly Labour ahead, if the North Lanarkshire and Livingston by-elections hadn’t been postponed until next year.

Anyway, that aside, basically everyone except Labour and Independents found themselves better off in percentage terms than in 2017, with the SNP growing the most. Though the Greens broadly did relatively well in these by-elections, having the second highest overall increase in vote share is partly down to standing in more of the wards than they did at the full elections.

Since two of the ten by-elections were Independent-dominated Islands wards, we can remove them from the total to get a clearer view of partisan shifts. Note that whilst Skye is, obviously, an island, the by-election there was well contested by parties, so it’s kept in the total:

Image Version

SNP - 12106 (37.6%, +5.7)
Conservative - 9730 (30.2%, +3.0)
Liberal Democrat - 3845 (12.0%, +1.9)
Labour - 2587 (8.0%, -5.0)
Green - 2271 (7.1%, +3.3)
Independent - 1510 (4.7%, -9.3)
Other - 121 (0.4%, +0.4)

This basically just has the effect of slightly increasing almost everyone else’s total vote share and swing versus 2017. The only exception is for Labour who drop slightly, as they were the only party to contest either of the Islands by-elections, with a creditable performance in Orkney’s North Isles.

The final measure would be to look only at the six wards where all of the Holyrood parties put in an apperance:

Image Version

SNP - 10334 (40.6%, +5.6)
Conservative - 7239 (28.4%, +0.1)
Liberal Democrat - 3147 (12.4%, -0.3)
Labour - 2367 (9.3%, -6.3)
Green - 1689 (6.6%, +2.2)
Independent - 599 (2.4%, -1.5)
Other - 76 (0.3%, +0.3)

The SNP come off the best here, but as with this whole post that’s largely a function of what wards these were. For example, had Labour contested Eilean a’ Cheò, that’d have brought a weak ward for them and the Conservatives into play whilst boosting the Lib Dems and Greens. Conversely, an absent Lib Dem in Craigentinny/Duddingston would have hurt Labour and Green shares. 

However, imperfect as all of these measures are for telling us about precise levels of support at a national level, they can act as useful corroboration of the general direction of support shifts. As they have been in most national recent polls, the SNP and Greens are up, the Lib Dems are relatively static, and Labour are worse off.

The Conservatives are perhaps the only outlier versus polling, coming out slightly improved or static whereas polling suggests a small decline. Of all the parties they are the one most likely to be benefitting from the low turnout of by-elections and the absence of Independents, which may explain it.

Though the early part of next year looks set to see plenty of ongoing pandemic disruptions to by-election schedules, I’m hopeful both on the basis of being a nerd and for saving lives that increasing rollout of a vaccine will bring more normality as the year goes on. There are already seven by-elections on the docket, alongside the Scottish Parliament and a UK Parliament By-Election as SNP MP for Airdrie and Shotts, Neil Gray, seeks to become MSP for… Airdrie and Shotts.

As usual, the second part of the Year in Review will look at Parliamentary polling. Look out for that tomorrow, as I take full advantage of my first day off my job to do… my other sort of job!

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)