Local Council by-elections 11th April 2024

[Tony Otim]

That is a really poor result for the SNP.

[certain]

Inverness South: Independent elected by 1247 to 1235 for LD at stage 7.

[olympian95]

Understand the LDs lost by 12 in the final round.

Dreadful result for the SNP.

[carlton43]

Apr 12, 2024 10:19:25 GMT The Bishop said:

Apr 12, 2024 9:56:24 GMT carlton43 said:

'A Very Good Result'  in which they actually lose the seat.

Please explain this to me?

Because the main local elections in Scotland use multi-member divisions elected by STV.

We had another example just last month when the Greens gained a seat from Labour, despite there actually being a significant swing *to* the latter from the 2022 result.

I understand the mechanics as I used to live in Scotland. 

My quibble is with the term 'very good result'.  Useful swing ... Yes, but expected between any leading party and the SNP at present.  Good result ...Hardly, as they lost the seat.  Very good result ... Utter nonsense.

But no matter.

[islington]

Apr 12, 2024 9:53:24 GMT certain said:

Whether or not a single-seat contest is described as “STV” or “AV” is not really of much importance. However, Counting Rule 55 specifies that a candidate is elected when they have achieved a majority of valid votes – simple AV. This makes the STV “quota” irrelevant in single-seat contests. Unfortunately, the bug-strewn computerised counting system doesn't know about Rule 55 and completely ignores it. The quota is calculated and published. If a candidate achieves an overall majority, but without reaching the quota, a further nonsensical counting stage is done wherein some runner-up's votes are transferred, by some mysterious process, to the winner.

Furthermore, the system does not include the final result details in the declaration – only the first preferences. The result of this is that the media usually fails to publish these important details, even sometimes who came second.

Can one of its advocates please remind us again of why this system is so much better than dear old FPTP?

[The Bishop]

Apr 12, 2024 10:49:13 GMT carlton43 said:

Apr 12, 2024 10:19:25 GMT The Bishop said:

Because the main local elections in Scotland use multi-member divisions elected by STV.

We had another example just last month when the Greens gained a seat from Labour, despite there actually being a significant swing *to* the latter from the 2022 result.

I understand the mechanics as I used to live in Scotland. 

I know, which is why your post puzzled me slightly.

Put it this way, the contest was almost universally expected to be a straight fight between the Independent and SNP. Hardly anybody predicted such a strong LibDem showing - on that basis at least it is indeed a good result for them.

But as you say, no matter. Its all about opinions at the end of the day

[Khunanup]

Link to the transfers:

www.highland.gov.uk/download/downloads/id/28590/candidate_votes_per_stage_report_ward-19.pdf

SNP continue to be transfer toxic (as are the Tories). They're also not keen on transferring to others either. It was their transfers that was critical to the final result though, after the Conservative transfers the Lib Dem was up by 15, the Indy got 27 more from the SNP than the Lib Dem so got over the top (the majority of the SNP vote was non-transferable however).

[carlton43]

Apr 12, 2024 10:21:09 GMT The Bishop said:

Anyway, the long SNP drought in Scottish council by-elections continues.

Yes.  That is a more relevant remark, because it looked to be the sort of seat they might have expected to win/hold.  But it is an odd seat large in territory with most of the electorate at the far north end in greater Inverness.  But the demographics here are changing all the time as different types of expansion take place drawing in different types of 'outsider' including a significant number of English and even foreign potential voters.

[dizz]

Apr 12, 2024 10:56:26 GMT islington said:

Apr 12, 2024 9:53:24 GMT certain said:

Whether or not a single-seat contest is described as “STV” or “AV” is not really of much importance. However, Counting Rule 55 specifies that a candidate is elected when they have achieved a majority of valid votes – simple AV. This makes the STV “quota” irrelevant in single-seat contests. Unfortunately, the bug-strewn computerised counting system doesn't know about Rule 55 and completely ignores it. The quota is calculated and published. If a candidate achieves an overall majority, but without reaching the quota, a further nonsensical counting stage is done wherein some runner-up's votes are transferred, by some mysterious process, to the winner.

Furthermore, the system does not include the final result details in the declaration – only the first preferences. The result of this is that the media usually fails to publish these important details, even sometimes who came second.

Can one of its advocates please remind us again of why this system is so much better than dear old FPTP?

You're not in a pub with us in Islington now Islington, so don't derail this thread!

[ricmk]

Question on my mind: The SNP local election results over the past year have been far far worse than Lib Dem local election results in 2014-5, and we can see further here how they are struggling on transfers.
We all know what happened to the Lib Dems in 2015. Why are we not projecting even worse for the SNP in these circumstances? I know Scotland-only polling has them falling back but not existentially. But local results, actual votes in the ballot box, paint a different picture. What are we missing?

[East Anglian Lefty]

Apr 12, 2024 10:56:26 GMT islington said:

Apr 12, 2024 9:53:24 GMT certain said:

Whether or not a single-seat contest is described as “STV” or “AV” is not really of much importance. However, Counting Rule 55 specifies that a candidate is elected when they have achieved a majority of valid votes – simple AV. This makes the STV “quota” irrelevant in single-seat contests. Unfortunately, the bug-strewn computerised counting system doesn't know about Rule 55 and completely ignores it. The quota is calculated and published. If a candidate achieves an overall majority, but without reaching the quota, a further nonsensical counting stage is done wherein some runner-up's votes are transferred, by some mysterious process, to the winner.

Furthermore, the system does not include the final result details in the declaration – only the first preferences. The result of this is that the media usually fails to publish these important details, even sometimes who came second.

Can one of its advocates please remind us again of why this system is so much better than dear old FPTP?

I'm not a great advocate for STV, but I don't think even its advocates would argue that it works particularly well for single-seat by-elections. The argument is that this is permissible because of the improvement for whole-council elections (or that you should use a different system for filling casual vacancies.)

[iain]

Redistributing the votes from the Independent gives the following result in Inverness South, with changes from 2022 (also redistributing Indy preferences at that election):

Lib Dem	856	25.9%	+12.0%
SNP	695	21.0%	-20.0%
Conservative	650	19.7%	-2.3%
Labour	447	13.5%	+3.8%
Green	272	8.2%	+1.4%
Alba	135	4.1%	+1.8%
Sovereignty	66	2.0%	New
Non-transferrable	184	5.6%	+1.3%

[islington]

Apr 12, 2024 11:01:34 GMT dizz said:

Apr 12, 2024 10:56:26 GMT islington said:

Can one of its advocates please remind us again of why this system is so much better than dear old FPTP?

You're not in a pub with us in Islington now Islington, so don't derail this thread!

I stand reproved.

Please take my question as being rhetorical in nature.

[Peter Wilkinson]

Apr 12, 2024 9:53:24 GMT certain said:

Whether or not a single-seat contest is described as “STV” or “AV” is not really of much importance. However, Counting Rule 55 specifies that a candidate is elected when they have achieved a majority of valid votes – simple AV. This makes the STV “quota” irrelevant in single-seat contests. Unfortunately, the bug-strewn computerised counting system doesn't know about Rule 55 and completely ignores it. The quota is calculated and published. If a candidate achieves an overall majority, but without reaching the quota, a further nonsensical counting stage is done wherein some runner-up's votes are transferred, by some mysterious process, to the winner.

Without having read the rules, I rather suspect that the "mysterious process" transfers the votes of the runner-up on the basis of whether they expressed any numerical preference whatever for the winner or just left a blank on the winner's line. But in any STV or AV election, unless one allows for one specific possible wrinkle (which I will take up again below), stating numerical preferences for all the listed candidates should produce exactly the same result, in terms of winners and losers, as doing so for all but one of them - so carrying out that final stage does seem pointless. Though, if every single voter had used all their preferences, that would mean that the eventual winner would be being reckoned to having been unanimously (and not just nem. con.) in the final count for a single seat.

Furthermore, the system does not include the final result details in the declaration – only the first preferences. The result of this is that the media usually fails to publish these important details, even sometimes who came second.

In fact, on the basis of just the first preferences, any of the six top candidates mathematically might have won - depending on the exact distribution of voters' lower preferences (though I would accept that any such distribution that produced a Green or even Labour victory in this case would come close enough to statistical impossibility in any genuine election as to be regarded with the gravest suspicion).

Time to bring in the above-mentioned wrinkle - what does Rule 55 mean by a valid vote? Or indeed a majority of them? If a vote remains valid even when all preferences stated on it have been eliminated and still needs to be accounted for in the determination of whether or not an absolute majority (not just a plurality) of valid votes have been cast for a particular candidate, then the final counting stage would have a purpose - because the result of any election in which no candidate gained such a majority would automatically be voided by that very fact. For instance, that would have applied to the result of this election, on the sole condition that no (or not enough) people had expressed any lower preferences.

However, that would be extremely peculiar in any rules governing STV elections - since in multi-member contests, these should allow the election of any candidates who have reached quota, even if votes outside that quota have expressed no preference for those candidates. Rule 55, therefore, is presumably a backstop to another rule, stating that candidates who have reached quota have been elected - in which case, Rule 55 would be useless, given my interpretation in the immediately previous paragraph, because in a single-seat election, the quota will have been determined as the minimum number of votes required as an absolute majority of valid votes on the first count (and in multi-seat elections, would be lower than this).

So, for Rule 55 to have any purpose, my interpretation must be wrong - it has to be there specifically to allow the election of candidates who have not reached quota but nevertheless have a majority of the votes that still have preferences for uneliminated (and so far unelected) candidates. It certainly looks as if the computerised counting system probably does elect the correct candidate(s) - but not by the correct rules, or including enough information in the declaration to verify that the candidates declared elected are indeed the correct ones.

[andrewp]

An update on local by election statistics this municipal year, following the two contests on 11/04/24

Excluding countermanded elections, up to 11 April 2024 there have now been 168 ordinary by elections for 170 seats since May 4th 2023 . In 2022/23 year there were 183 by elections, in 2023/24, there will be 175, so a slight decrease

The Conservatives have defended 47- Held 19 and lost 28: 5 to the Greens, 16 to the Lib Dems, 6 to Labour and 1 to an Independent. ( retention rate 40%) and have gained 10

Labour have defended 52- Held 37 and lost 15- 5 to the Conservatives, 5 to Independents, 3 to the Lib Dems and 2 to the Greens( retention rate 71%) and have gained 12

Lib Dems have defended 33 - Held 29 and lost 4, 1 each to the Conservatives and Greens and 2 to Independents ( retention rate 88%) and have gained 23

Greens have defended 10- Held 6 and lost 4: 1 each to the Conservatives, Labour, the Liberal Democrats and to an Independent( retention rate 60%) and have gained 10

There have been elections for 14 seats previously won by Independents: 8 have been won by an another Independent , 2 have been lost to the Greens and 1 each to the Conservatives, Labour, the Lib Dems and Plaid Cymru.And Independents have gained 10
Residents/ local groups have defended 6, held 4 and lost 2 to the Lib Dems.
SNP have defended 5, lost 5, 2 to the Conservatives and 3 to Labour ( retention rate 0%)
PC have defended 3, held 1 , lost 2, 1 to Labour and 1 to an Independent (retention rate 33%) and have gained 1

Overall net changes

Con -18
Lab -3
LD +19
Green +6
Ind + 4
SNP -5
PC -1
Res/ local groups -2

[timrollpickering]

Apr 12, 2024 10:57:51 GMT Khunanup said:

SNP continue to be transfer toxic (as are the Tories). They're also not keen on transferring to others either. It was their transfers that was critical to the final result though, after the Conservative transfers the Lib Dem was up by 15, the Indy got 27 more from the SNP than the Lib Dem so got over the top (the majority of the SNP vote was non-transferable however).

Do we know the SNP voters plumped en mass or did many transfer to the already elminated Alba and Greens?

[The Bishop]

Apr 12, 2024 9:40:34 GMT andrewp said:

I make that

Ind 22.1 % ( new Indy)
LD 19.7% (+11.1)
SNP 19.4% (-15.8)
Con 16.1% (-0.8)
Lab 11% (-3.7)
Green 7.2% (+1.8)
Alba 3.2% (+1.3)
Sov 1.2% ( new)

Labour are up by that amount I think, not down?

(not that it really matters, save perhaps for the jamesdoyle GWBW calculations)

[andrewp]

Apr 12, 2024 12:56:23 GMT The Bishop said:

Apr 12, 2024 9:40:34 GMT andrewp said:

I make that

Ind 22.1 % ( new Indy)
LD 19.7% (+11.1)
SNP 19.4% (-15.8)
Con 16.1% (-0.8)
Lab 11% (-3.7)
Green 7.2% (+1.8)
Alba 3.2% (+1.3)
Sov 1.2% ( new)

Labour are up by that amount I think, not down?

(not that it really matters, save perhaps for the jamesdoyle GWBW calculations)

Yes, you are correct. I will amend.

[iain]

Apr 12, 2024 12:42:01 GMT timrollpickering said:

Apr 12, 2024 10:57:51 GMT Khunanup said:

SNP continue to be transfer toxic (as are the Tories). They're also not keen on transferring to others either. It was their transfers that was critical to the final result though, after the Conservative transfers the Lib Dem was up by 15, the Indy got 27 more from the SNP than the Lib Dem so got over the top (the majority of the SNP vote was non-transferable however).

Do we know the SNP voters plumped en mass or did many transfer to the already elminated Alba and Greens?

Many transferred to Alba / the Greens. SNP voters weren't more likely not to transfer than anyone else.

[johnloony]

Apr 12, 2024 11:57:41 GMT Peter Wilkinson said:

Apr 12, 2024 9:53:24 GMT certain said:

Whether or not a single-seat contest is described as “STV” or “AV” is not really of much importance. However, Counting Rule 55 specifies that a candidate is elected when they have achieved a majority of valid votes – simple AV. This makes the STV “quota” irrelevant in single-seat contests. Unfortunately, the bug-strewn computerised counting system doesn't know about Rule 55 and completely ignores it. The quota is calculated and published. If a candidate achieves an overall majority, but without reaching the quota, a further nonsensical counting stage is done wherein some runner-up's votes are transferred, by some mysterious process, to the winner.

Without having read the rules, I rather suspect that the "mysterious process" transfers the votes of the runner-up on the basis of whether they expressed any numerical preference whatever for the winner or just left a blank on the winner's line. But in any STV or AV election, unless one allows for one specific possible wrinkle (which I will take up again below), stating numerical preferences for all the listed candidates should produce exactly the same result, in terms of winners and losers, as doing so for all but one of them - so carrying out that final stage does seem pointless. Yes it is. Though, if every single voter had used all their preferences, that would mean that the eventual winner would be being reckoned to having been unanimously (and not just nem. con.) in the final count for a single seat. Yes it would.

Furthermore, the system does not include the final result details in the declaration – only the first preferences. The result of this is that the media usually fails to publish these important details, even sometimes who came second.

In fact, on the basis of just the first preferences, any of the six top candidates mathematically might have won - depending on the exact distribution of voters' lower preferences (though I would accept that any such distribution that produced a Green or even Labour victory in this case would come close enough to statistical impossibility in any genuine election as to be regarded with the gravest suspicion). When AV elections are done properly, each round eliminates as many candidates as possible - i.e. any which have no chance of winning (e.g. if the bottom two candidates have fewer votes than the next lowest). Unfortunately Scotland doesn’t do AV properly.

Time to bring in the above-mentioned wrinkle - what does Rule 55 mean by a valid vote? A valid vote is a ballot paper on which there is still a preference expressed for a continuing candidate (i.e. any which hasn’t gone into the “non-transferable” pile). Or indeed a majority of them? If a vote remains valid even when all preferences stated on it have been eliminated and still needs to be accounted for in the determination of whether or not an absolute majority (not just a plurality) of valid votes have been cast for a particular candidate, then the final counting stage would have a purpose - because the result of any election in which no candidate gained such a majority would automatically be voided by that very fact. For instance, that would have applied to the result of this election, on the sole condition that no (or not enough) people had expressed any lower preferences. That would be logical if the non-transferable pile were a “Re-Open Nominations” option, which it isn’t.

However, that would be extremely peculiar in any rules governing STV elections - since in multi-member contests, these should allow the election of any candidates who have reached quota, even if votes outside that quota have expressed no preference for those candidates. Rule 55, therefore, is presumably a backstop to another rule, stating that candidates who have reached quota have been elected - in which case, Rule 55 would be useless, given my interpretation in the immediately previous paragraph, because in a single-seat election, the quota will have been determined as the minimum number of votes required as an absolute majority of valid votes on the first count (and in multi-seat elections, would be lower than this).

So, for Rule 55 to have any purpose, my interpretation must be wrong - it has to be there specifically to allow the election of candidates who have not reached quota but nevertheless have a majority of the votes that still have preferences for uneliminated (and so far unelected) candidates. Yes it is. It certainly looks as if the computerised counting system probably does elect the correct candidate(s) - but not by the correct rules, or including enough information in the declaration to verify that the candidates declared elected are indeed the correct ones.

In AV, and indeed in STV, there are effectively two different quotas. The original quota (what I call the large quota) is the total votes divided by (n+1). The continuing quota (what I call the small quota) is the quota, recalculated at each round, to allow for the number of non-transferable votes. The small quota gradually gets more and more different from the large quota, depending on how many non-transferable votes there are.

The problem is that virtually nobody else (except me) refers to the small quota as a “quota”. It is defined properly in the rules (and in the legislation) but those rules are not fully understood by some people who don’t read them properly. That includes the booliaks whomwrote the computer programme which counts the votes in AV elections and which does the redundant final round.