North East (Dundee)
Dundee’s North East by-election was one I had down as a meaningful yet dull contest. Coming about in unfortunate circumstances with the death of Labour councillor Brian Gordon, this was a ward where the SNP had never won less than 50% of the vote, and a council in which the SNP were only one seat short of a majority. I expected an easy victory to hand them that majority.
Sure enough the SNP did win the seat and clinch that majority, making Dundee the only single-party majority council in Scotland. However this also saw them drop below 50% for the first time, and the count kept going until every other candidate was eliminated. First preferences in full;
- SNP – 1507 (46.9%, -6.9)
- Labour – 1224 (38.1%, +11.1)
- Conservative – 271 (8.4%, -0.7)
- Anti-Cuts Coalition – 91 (2.8%, +1.5)
- Green – 77 (2.4%, +0.8)
- Citizens First – 45 (1.4%, +1.4)
Although the SNP dropped below 50% it wasn’t by much. Labour meanwhile managed a quite impressive increase in their vote share, compared to the Conservatives on a second of their relatively rare by-election declines in a row. On the smaller party front both the Anti-Cuts Coalition, which is a brand of (the largely defunct) TUSC, and the Greens also saw gains, though from very low bases. Note that although the Lib Dems are missing here, they only won 1.3% last time, so that’s not a lot of votes to redistribute. There was also a reasonably well performing Independent in 2017 with 5.8% whose votes were up for grabs.
Comparing the final head-to-head round at stage 5 with the equivalent last time;
- SNP – 1576 (49.0%, -6.7)
- Labour – 1388 (43.2%, +8.1)
- Didn’t Transfer – 251 (7.8%, -1.3)
Labour saw a remarkable narrowing of the vote when it came down to just them and the SNP, but not enough to take the seat. If this result was replicated in 2022 though the ward would return to the 2 SNP, 1 Labour balance that has been long established.
It’s worth some general notes here about the awkwardness of by-elections, of which this is a particularly dramatic (albeit rare) example. Single member by-elections for wards that normally elect multiple members via STV can result in some voters losing their voice. Here, despite a decrease in the vote for the SNP and substantial increase for Labour, the SNP have ended up with every seat in the ward, since only one seat was up. Most voters are now not represented by a councillor of their most preferred party, something that isn’t meant to happen in Scotland.
You can wave that away with claims of councillors being elected personally and not just as anonymous party bodies, but honestly that claim is almost always overblown. Most voters even at local level still vote primarily on party lines. A good candidate or a long-serving councillor can make quite a lot of difference, but they’ll still mostly get in because they have the right logo next to them on the ballot. By-elections can unbalance that.
Additionally, by-elections can empower a small minority of voters to effect a change to the entire council. In this case that’s been giving the SNP a majority on the council that they didn’t win in 2017 when the entire city was able to vote. Democratically speaking that isn’t a particularly fair advantage to give to just one wards worth of voters.
As much as I enjoy covering council by-elections, they are in some respects a sign of a flawed electoral system. The UK as a whole has such a strong notion that representatives are elected personally, even though that is overwhelmingly untrue, that it’s hard to envision an alternative arrangement (for example co-option, where the party of the outgoing councillor selects a replacement) being accepted by the general public. But nonetheless, it’s something we should be aware of.
Anyway, since this was another machine count, there’s the full set of data available for this by-election that you don’t get with a hand count. The most interesting thing to look at here would be where each party’s second preferences went, which isn’t fully captured in the normal process of counting and elimination.
As with Clackmannanshire Central at the end of March, the supporters of the three big parties proved very reluctant to mark any later preferences, with bullet votes making up more than 40% of each party’s total. Without a Kipper on the ballot this time it was the Conservatives who proved the least transfer friendly for everyone else, though oddly they were only joint last in terms of Green later preferences alongside the Anti-Cuts Coalition. Speaking of the Anti-Cuts Coalition, voters for the rather odd Citizens First candidate were very keen to transfer to them. Could be people just voting for the non-mainstream options?
The SNP, Greens and Labour were all relatively transfer friendly amongst themselves, whilst the Conservatives proved most likely to give Labour their second preference this time. I’d guess that had there been a Lib Dem available they would have taken a good chunk of those Conservative second prefs but couldn’t say for sure.
Haddington and Lammermuir (East Lothian)
Down in East Lothian, Haddington and Lammermuir was a relative rarity in that in 2017 the SNP hadn’t been in the top two. With East Lothian as a whole being one of Labour’s last remaining areas of strength in Scotland, their chances looked pretty good.
On the day however they slipped into third place, and it was the Conservatives that took the seat. I had noted that there wasn’t a lot between the SNP and Labour last time, and that if it came down to an SNP-Conservative head-to-head instead that it’d likely go in the latter’s favour. I was not wrong, as it turned out. First preferences in full;
- Conservative – 2212 (35.0%, +6)
- SNP – 1866 (29.5%, +3.5)
- Labour – 1359 (21.5%, -12.2)
- Lib Dem – 774 (12.2%, +5)
- UKIP – 108 (1.7%, +1.7)
In the complete opposite of the previous week’s by-election (don’t be so surprised, that’s the council by-elections game), Labour’s vote plummeted to the gain of everyone else – remember too that the Greens weren’t standing this time, leaving their 4% going spare as well. After the Conservatives, the Lib Dems were the biggest beneficiaries, picking up just a little bit of the energy that propelled them to success in England the week before. When we look at the final head-to-head at stage 4;
- Conservative – 2759 (43.7%, +10.0)
- SNP – 2469 (39.1%, +39.1%)
- Didn’t Transfer – 1091 (17.3%, -4.4%)
Although the Conservative lead held throughout the count, it narrowed a bit by the end. It was also a fair bit closer, at 4.6%, than the equivalent lead that Labour held over the Conservatives in the 2017 recalculation, which was 10.9%. Whilst I was right in saying the Conservatives would win a Con-SNP head-to-head, the SNP actually received more Labour transfers than the Conservatives this time round, which likely helped close that gap.
It’s quite useful though to have two starkly different by-election results like this a week apart and covered in the same post. It means I can again make the point I keep desperately trying to drum into people; Scotland is a politically diverse country and if you try and hang some grand theory of how national politics is going on individual council by-elections, you’re on to plums.
We’re being absolutely spoiled with loads of machine counts this year, meaning there’s more of that juicy data we can use to look at second preferences for each party;
Once again the rate of bullet voting was huge for everyone but the Lib Dems, who also proved the most popular destination for second prefs from the big three. They themselves tilted a bit Labour but with a relatively even split for the SNP and Conservatives as well, whilst UKIP plumped for the Conservatives.
And that’s us for 2019’s by-elections so far! There aren’t any more scheduled at the moment, but it’s quite likely we’ll see at least a couple more by the end of the year.