Fife and Inverness Council By-Election Results, 14/11/19

November’s flurry of council by-elections hit its peak this week, with three by-elections to keep on top of. Unlike last week’s Shetland by-elections all of these were heavily partisan affairs, with all five of the Holyrood parties taking part. With a General Election looming, I remind readers not to hook too many theories about the national situation on the shoogly peg of a handful of local by-elections.

Fife council followed in Shetland’s footsteps by holding two by-elections at the same time in neighbouring wards. We’ll start with Dunfermline Central since it was first to declare. This was to my mind classic STV, being caused by a Conservative resignation, in a ward the SNP led in first preferences in 2017, but where Labour would have won a single councillor election. Beautifully messy! I reckoned it was far too close between those three parties for me to call.

I was absolutely right to say it was going to be close, but with a party I hadn’t really thought of. Although I’d noted that the Lib Dems were formerly strong in this ward, I foolishly hadn’t considered that their recent revival might be most evident here. Looking at first preferences in full;

  • SNP – 1526 (33.2%, +3.4)
  • Conservative – 1142 (24.8%,+0.7)
  • Liberal Democrat – 1050 (22.8%, +15.9)
  • Labour – 612 (13.5%, -13.1)
  • Green – 235 (5.1%, +1.9)
  • Libertarian – 28 (0.6%, +0.6)
  • Note: 2 Independent candidates won 9.3% between them in 2017.

Good news for most parties here, as they all increased their share versus the last election – that will have been aided by the fact neither of the Independents from then opted to stand in the by-election. It’s the Lib Dems who had the most notable increase however, more than trebling their share with a whopping 15.9% increase. Not only did that move them from fifth to third place in terms of first preferences, they ran the Conservatives pretty close for second.

The only party to lose out were Labour, whose vote share was cut in half by that nasty -13.1%. As noted in the preview post, Labour may well have been boosted in 2017 by one of their candidates being former Pars player and manager Jim Leishman, but his absence alone can’t account for that level of swing. Now, the even more dramatic final head-to-head round at stage 5;

  • SNP – 1798 (39.1%, +5.4)
  • Liberal Democrat – 1796 (39.0, +39.0)
  • Didn’t Transfer – 1008 (21.9%, -3.3)
  • Note: Labour were in the 2017 head-to-head with 41.1%

Nerve-wracking déjà vu for the Fife SNP and Lib Dems after transfers here here. It came down to just two votes between them, a la North East Fife at the 2017 general election. You don’t get much more exciting than this in by-elections!

Nipping down the road to Rosyth, I had this down as being much more straightforward and likely to go to the SNP, who had caused the original vacancy.

That proved to be the right call, and Rosyth accordingly gave us the least dramatic results of the night. First preferences in full;

  • SNP – 1347 (42.8%, +6.3)
  • Conservative – 768 (24.4%, +2.1)
  • Labour – 480 (15.2%, -3.0)
  • Liberal Democrat – 249 (7.9%, +0.5)
  • Independent – 157 (5.0%)
  • Green – 132 (4.2%, +1.3)
  • Libertarian – 16 (0.5%, +0.5)
  • Note: 3 Independent candidates won 12.7% between them in 2017.

A bit like in Dunfermline, this saw gains for every party except for Labour – again, note that there were 3 Independents here in 2017, none of whom returned this time. Overall movement was much more modest though, with Labour only down 3% and the Lib Dems barely up by 0.5%. The SNP achieved the largest gain here, up 6.3% which helped give them a completely unassailable lead. So looking at the final round at stage 6;

  • SNP – 1639 (52.0%, +9.9)
  • Con – 976 (31.0%, -2.6)
  • Didn’t Transfer – 534 (17.0%, -7.3)

After transfers the SNP actually achieved an even larger swing, widening a gap of 8.5% between them and the Conservatives at the equivalent stage in 2017 to 21% here. Crossing 50% also gave them a relatively rare example of a machine count not needing to go all the way to eliminating every other candidate.

Finally, let’s head up north for the result of Highland Council’s Inverness Central by-election. This was similar to Rosyth in that it was an SNP resignation that triggered it, and I thought it likely they’d come out on top. (Also note the slight shifting of the position of the key in this chart, having had time to learn the lesson between the Dunfermline Central count and this one…)

Again, that call was absolutely spot on, as the SNP very comfortably won this one. First preferences in full;

  • SNP – 1015 (45.2%, +12.3)
  • Conservative – 345 (15.3%, +2.9)
  • Independent – 277 (12.3%)
  • Liberal Democrat – 237 (10.5%, +6.0)
  • Green – 220 (9.8%, +3.7)
  • Labour – 154 (6.9%, -10.0)
  • Note: 3 Independent candidates won 27.3% between them in 2017, and one of them was elected as a councillor.

Even more than in the other two wards in this post, the caveat about Independent candidates is crucial here. In 2017, Janet Campbell won 15.3% of the vote and was elected as a councillor and so obviously would not re-appear in the by-election. That will account for some portion of the SNP’s big gain of 12.3%, as well as the gains made once again by three of the other parties. On the other hand, Labour’s result might have looked even worse if the three 2017 independents had been on the paper. At that election Labour won the other one of this ward’s three councillors, so shedding 10% of their vote and coming last is a pretty grim showing for them.

Votes at the final round in stage 3 (changes versus the “pure” 2017 recalc);

  • SNP – 1115 (49.6%, +4.4)
  • Conservative – 360 (16.0%, +16.0)
  • Independent – 338 (15.0%)
  • Liberal Democrat – 325 (14.5%, +14.5)
  • Didn’t Transfer – 110 (4.9%, -15.2)
  • Note: Janet Campbell was in the 2017 head-to-head with 34.6%

Notice how not only are there three other candidates still in the count at the point at which the SNP were deemed elected, but also that the SNP hadn’t reached quota? Whereas most by-elections are machine counted and thus the software quickly and easily follows a strict “complete” STV process, this was a hand count where a number of time saving and common sense rules apply. With 1115 votes, the SNP have more than the 1023 votes sitting with the other three candidates by this stage. Since it was therefore impossible for any other candidate to achieve enough transfers to win, the SNP are elected and the count process finishes. That’s a neat little quirk of our electoral law, but hand counts in general do then deny us the other juicy data a machine count delivers.

Next week will see the the last two by-elections not just of this busy November but of 2019, then it’s a clear run until General Election day on the 12th!