Term in Review: 2017-22 By-Elections


Unbelievable as it may seem, it has indeed been almost five years since the last Scottish local elections in May 2017. Politics just has not stopped for even a single moment since then. In vote terms we’ve had a staggering four parliamentary elections in that time. First was the snap UK Parliament election in 2017, then unexpected final EU elections in 2019, another snap UK election later that same year, and finally a mercifully on-schedule Scottish Parliament election in 2021. Alongside those, the UK completed its formal departure from the EU after years of high drama, and there’s of course been the all-consuming pandemic.

If readers will indulge me with some personal reflections, it’s particularly hard for me to fathom that it really has been five years because 2017 was such a big deal for me at the time. I’ve made no secret of the fact I myself stood in those elections, as the Green candidate for Glasgow’s Victoria Park ward. I lost by just shy of 40 votes in the final round of transfers – to a Labour candidate I’m very fond of as an individual, so I couldn’t even bear a grudge!

Losing isn’t very fun at the best of times, especially not when it was so close and I’d put my all into the campaign. But it was especially painful for me as it came a few months after I had been made redundant, and I was therefore fully pitched into a period of very grim unemployment. Every cloud has a silver lining however, and it was both the fact I wasn’t an elected representative and had loads of spare time and energy that led to me setting up Ballot Box Scotland in January 2018.

BBS has been an absolute joy, and something that I love doing, even now that I am very happily employed again. I’d rather not have had that period of unemployment of course, but on honest reflection, I’m much happier being the elections nerd running this project than I would have been as a councillor. I’m surprised and delighted by just how much BBS has grown, and how much people seem to enjoy and appreciate it, and I’m looking forward to engaging with 2022’s elections in a completely different way to last time around.

I've got a lovely bunch of by-elections

Across the past five years, 65 (5.3%) of the 1227 councillors elected in May 2017 have been replaced following a total of 64 by-elections – one of these was a rare two-in-one. As three wards had two separate by-elections through the period, that means 61 (17.2%) of Scotland’s 354 council wards, which accounted for 16.6% of voters in 2017, went back to the polls at some point this term. They did so in 24 (75%) of the 32 council areas. You can get a visual sense of how these were spread throughout the country in this map:

If you wanted to, you can go right through the archives to find my coverage of every single one of these – including after-the-fact coverage of those from 2017, before BBS launched. Note that as you go further back in time, you’ll see a huge difference between what I produce now and what I did to begin with. Although I like to think this project helps improve other people’s understanding of elections in Scotland, it’s also been educational for me too! I can do a lot more with election data, in much prettier ways, than way back in 2018.

As always when talking about STV, we have to recognise how different it is from First Past the Post. The vacating party is not the same as the party that “won” overall at the previous election, which complicates reading shifts in voter support simply from shifts in seat numbers. Added to that mix are the large number of wards which elected Independent councillors in 2017, who for obvious reasons don’t appear on by-election ballots, which can then also limit how useful even the raw vote shifts are.

Unlike an annual review piece, I will not be summing up the total of those vote shares in this article anyway. Doing that across a whole year is already a bit of a stretch, as we all know a few days never mind a few months can lead to dramatic changes in support. Doing it across five years would be utterly pointless. By-elections in 2017, 2019 and 2021 might as well have taken place on different planets for all the difference in political context between each of those years. Equating them as one set of “changes vs 2017” in vote share wouldn’t really add to our understanding.

Seat Shifts

I’ll therefore just be looking at changes in seats. These still have the same caution that the political context of different years is incomparable, but at least have the tangible impact of a bum on a seat in a council chamber. The table below shows each of the by elections across the term with both the vacating party and 2017’s notional winner shown, alongside the eventual winner of the by-election.

And then this little chart shows those shifts more concisely. As with the table, it’s showing the difference between seats vacated (what councillors stood down) versus seats defending (who was the 2017 lead party in the relevant wards).

If we then compare in terms of vacated seats, which is more impactful on political control but not particularly useful for political understanding (note Independents use slightly different terminology as they aren’t directly comparable to one another):

Conservative - Vacated 15, gained 10, held 9, lost 6 (net +4)
SNP - Vacated 21, gained 11, held 12, lost 9 (net +2)
Liberal Democrat - Vacated 4, gained 4, lost 4 (net nc)
Labour - Vacated 12, gained 4, held 5, lost 7 (net -3)
Independent - Vacated 13, won 10 (net -3)

The net changes here are positive for the Conservatives and SNP, neutral for the Lib Dems, and negative for Labour and Independents. The fact that vacating and 2017 winning party are not the same is demonstrated by what look like very high numbers of losses by this measure for every party, with similarly high gains for most of them. In fact the Lib Dems even appear to have made a complete switch, losing in every ward they vacated, but gaining as many elsewhere.

Per my usual warning, whilst the above is important in terms of actual political composition of the council, it’s not a true measure of electoral shifts due to the way STV works. In addition, some of the Independent vacancies were councillors who had originally been elected for a party. We should therefore also compare with who would have been expected to win a single seat election in 2017.

SNP - Defending 17, gained 10, held 13, lost 4 (net +6)
Conservative - Defending 17, gained 5, held 14, lost 3 (net +2)
Liberal Democrat - Defending 3, gained 2, held 2, lost 1 (net +1)
Labour - Defending 13, gained 3, held 6, lost 7 (net -4)
Independent - Won 15 in 2017, won 10 through period (net -5)

By this more useful measure, the Lib Dems join the SNP and Conservatives in having a net gain, though the SNP now have a substantial lead in terms of the size of those gains. Labour and Independents come off even worse, though of course remember what would be the winning Independents from 2017 were not on the ballots, either due to still serving as councillors or (more sadly) their passing prompting the vote in the first place.

Apart from Labour, who had a particularly hard time of it over the period, figures for lost seats here are much milder than the other measure. In the majority of cases (35 by-elections), the same party that would have won a single seat vote in 2017 won the by-election. When it came to seats that were genuine electoral gains, most SNP gains came from Labour, and all of Labour’s gains came from the SNP. The Conservatives were more evenly mixed, and both Lib Dem gains came from Independents and therefore contests that were much more open.

A notable absence from any of this discussion has been the Greens. No by-election was prompted by a Green vacancy, nor did they win any, though they came close in Leith Walk. Even if they had caused a vacancy, there are very few wards they were 2017 winners in – just a handful in Edinburgh. However, a Green councillor did resign from one of those wards days after the 6-month rule preventing by-elections kicked in. That would have been a uniquely fascinating contest to watch.

Wrapping up

To round out this piece and tie a neat little bow on the entire 2017 to 2022 term, let’s end with something light-hearted…

The By-Election Bonanza Awards for most by-elections held go to Highland (for total number) and Clackmannanshire (for proportion). Whereas Highland were the clear leaders in total number of by-elections with 8, the 3 held in Clackmannanshire amounted to 60% of the local wards there.

The Serial Shufflers Award for most net change in councillor affiliation goes to South Lanarkshire. 10 of 64 councillors there, or 15.6% of the total, have a different affiliation compared to those elected in 2017.

South Lanarkshire also win the related That’s Not What We Voted For Award, as none of these changes came about from by-elections. (Long-time BBS followers may know I’ve raised democratic queries about by-elections in PR circumstances anyway, given the potential for distortion, but that’s another issue entirely!)

The Perfect Snapshot Award goes to East Ayrshire, the only council not to have a single by-election nor a (permanent?) change in any councillor’s political affiliation for the entire term. All 32 councillors are the same people, for the same parties, as elected in 2017.

Finally, the Most Enjoyable By-Elections Award goes, absolutely hands down, no competition whatsoever, to Highland Council. The transparency of Highland’s by-election counts has been superb. Pre-pandemic, when they did hand counts, they tweeted the results of each round of counting as and when they completed, which made for oddly thrilling drama. With machine counts in the pandemic they started providing a live broadcast of the count process and declaration. Okay, sound failing at the declaration for the first of these wasn’t great, but the dramatic music added a welcome air of camp to proceedings.

Now, that’s well and truly it for this sitting of Scotland’s councils. There are still 14 weeks to go until the vote, and we may still see some last minute resignations and defections from parties, but I’ll no longer be tracking those given they’ll have effectively zero impact. 

If you haven’t already and can afford to do so, you can chip in a little bit to support my coverage of the upcoming elections at the link below. I’ll be closing this crowdfunder shortly, though normal donation routes will remain open. I’ll soon be getting underway with a whole suite of coverage including party profiles, key wards to look out for, and more. It’s going to be another busy few months…

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)