top of page
Tom Monto

Scottish local authorities elections 2007 --- STV in practice

Updated: Feb 17

2007 Scotland local authority results


the following stats talk about votes received on 1st Count, not the number of votes that a candidate held when he or she was elected. In Scotland in 2007, successful candidates received range of percentage of votes in 1st Count, all of these are First preference votes - ranging from 50 percent (before surplus votes transferred away) to as low as 3 percent (before vote transfers received) so obviously below quota there was a grey zone where a set percentage (ranging from 3 to quota (20 or 25 percent) may be enough as a start for later incoming transfers to get the candidate be elected but not necessarily. Only one candidate was elected who had started out with as few as 3 percent - likely there were many, many who got more than 3 percent on the 1st Count who were not elected. for one thing, with low initial count, it is likely such a low-ranking candidate will be eliminated as the field of candidates is thinned through eliminations if the candidate does hang on til the end, it is likely an initially-less-popular candidate will not have the lead in votes over each other remaining candidate still in the running to take a seat. But I would say, no candidate can win with just 3 percent of the votes.

(front runners are the candidates in top three in 3-seat district or top four in 4-seat district, usually mentioned in regards the standings in the 1st Count

Things to notice Most front runners were elected in the end. Vote transfers seldom changed more than one front runner. Usually no change at all. Single voting in MMD already created balanced, mixed rep. Never was someone less popular elected over someone more popular, popularity being measured at that exact moment. This is why the first thing holds true - usually the ones most popular at the beginning remain the popular choice and thus are elected. Election can come about through attaining the quota or by being most popular at the end when plurality is used to fill last seat, or occasionally the last two seats. Parties that the candidate belonged to may be identified below but only for identification purpose. Parties play no part in the vote counting. Parties know they will not take many seats so most (or many?) run just one candidate in a district. Example Hazelhead ward (Aberdeen City) only one party ran more than one candidate. and no party took more than one seat Peak votes of successful candidate - the number of votes each has when elected - is usually about the same (around 1000). Where the peak vote is higher than that, you know that the candidate's votes were reduced to quota if still seats(s) remained to be filled.

6.9-FP-percent success Counting First preference votes, we see in the Aberdeen district Torry ward that a candidate was elected after getting just 6.9 percent of votes placed in the 1st Count. How she got enough votes to be elected - she eventually got 950, just shy of quota - is detailed below.


Apparently in the 2007 Scotland local authorities elections, a candidate who received just 3 percent FP-votes was successful. I am not sure where that person ran but it must be somewhere in the many wards in one of the districts. No doubt the candidate received more than those 3 percent votes by the time he or she was elected.

=============

More general statements and detail on the over all local authorities elections provided below.

As well, that is where you will find some discussion on how votes are transferred and the effects of such. ------------------

Aberdeen City 2007


looking at specific ward contests in Aberdeen City district , we see Dyce 4 seats 17.8 lowest First Preference percentage of successful candidate Bridge 4 seats 14.3 lowest First Preference percentage of successful candidate Kingshills 3 seats 13.6 lowest First Preference percentage of successful candidate Northfield 3 seats 10.1 lowest First Preference percentage of successful candidate. This was SNP. SNP ran two candidates and one received 43 percent of FP votes, the other 10.2. the lesser received 500 vote transfers from his party mate's surplus and 100 from elsewhere to win seat with just more than quota. SNP's leader's surplus went mostly to the other SNP candidate but ten percent of them went to Aberdeen Labour and every other candidate got some as well, showing how voters have liberty to rank preferences how they want, sometimes even apart from party lines. 1 "Turn-over" happened in Northfield -- a candidate with 10.2 First Preference percentage did not win. This was Conservative. the Conservative party ran one candidate but received only 30 vote transfers from others and never took quota. so 10.1 percent of First Preference votes was enough to be base for success while more than that was shown to be not enough sometimes. Hilton 3 seats 11.9 lowest First Preference percentage of successful candidate "Turn-over" 16.8-FP percent front runner not elected in the end. Tillydrone 3 seats 8.5 lowest First Preference percentage of successful candidate "Turn-over" 10.7-FP percent front runner not elected in the end. Midstocket 3 seats front runners ranged from 21 to 27 First Preference percentage all front runners elected in the end with quota. George Street 4 seats 11.4 lowest First Preference percentage of successful candidate "Turn-over" 13.8-FP percent front runner not elected in the end. Lower Deeside 3 seats front runners ranged from 24 to 17 First Preference percentage all front runners elected in the end with quota. Hazelhead 4 seats 13.2 lowest First Preference percentage of successful candidate all front runners elected in the end with quota. quota 1548 peak votes 1967, 1618, 1650, 2238 of successful candidates (due to after-success transfers of surplus votes, peak votes may overlap.) Airyhall 3 seats 21.1 lowest First Preference percentage of successful candidate all front runners elected in the end, one without quota. Torry 4 seats 6.9 lowest First Preference percentage of successful candidate "Turn-over" 7.8-FP percent front runner not elected in the end. quota 1000 peak votes 1646, 1023, 1138, 950 of successful candidates (due to after-success transfers of surplus votes, peak votes may overlap.) together the two SNP candidates took 40 percent of vote in the 1st Count (2000 votes) so by that, party was eligible for 2 seats and it did win two seats. 6.9-percent FP candidate (SNP) was assisted by receiving vote transfers from the other SNP candidate after his election. (this infusion meant that the lesser SNP candidate was out of reach of eliminations -- he had risen to 4th place (including the successful SNP candidate). the lesser SNP candidate was safe in the top four but also luckily for the lesser SNP candidate, there were five candidards of various parties or no parties even less popular than him - including two Independents and a Libertarian - they were used up in the elimination stages shielding the lesser SNP candidate from elimination, while also a good portion of their transfers went to the remaining (lesser) SNP candidate. Kincorth 4 seats 15.2 lowest First Preference percentage of successful candidate all front runners elected in the end, all with quota. quota 960 peak votes of successful candidates range: 1465 to 966 (due to after-success transfers of surplus votes, peak votes may overlap.) ================================

Fife district although the Fife district is divided into 23 wards electing three or four members each, still consistency and relative ranking was tight. no party took more seats than a more popular party.

each party took a seat for about same number of voters Labour 24 seats 40,000 votes 1700 per seat SNP 23 seats 39,000 1700 per seat Liberal Democrats 21 seats 31,000 1500 per seat Conservative 5 seats 15,000 3000 per seat of course, party that takes less seats will have more wasted because votes do not leave wards or districts (same as under FPTP's districts or wards) in district round-up we see Scottish Green took 1900 votes and no seats - it is not stated over how many of the district's 23 wards, those 1900 votes were spread. (If they were spread over multiple candidates in the same ward, then vote transfers should have done the party some good.) Buckhaven has record for the succ. candidate with the least number of First preference votes. only four times in Fife district was a front runner in the 1st Count not elected in the end.

Fife district Ward results West Fife 4 seats front runners elected 15.6 lowest percent of FP votes taken by succ. cand Dumferline North 3 seats front runners elected 23.4 lowest percent of FP votes taken by succ. cand Dumferline Central 4 seats front runners elected 19 lowest percent of FP votes taken by succ. cand Dumferline South 4 seats 9 lowest percent of FP votes taken by succ. cand 1 "turn-over" 10.6-percent FP-vote cand not elected, 9-percent FP-vote cand elected Rosyth 3 seats front runners elected 20.8 lowest percent of FP votes taken by succ. cand Inverkeithing 4 seats front runners elected 17.5 lowest percent of FP votes taken by succ. cand Lochs 3 seats front runners elected 20.8 lowest percent of FP votes taken by succ. cand Cowdonbeath 3 seats front runners elected 18.9 lowest percent of FP votes taken by succ. cand Lochgelly 3 seats front runners elected 23.4 lowest percent of FP votes taken by succ. cand Burtisland 3 seats front runners elected 15.6 lowest percent of FP votes taken by succ. cand Kirkcaldy North 3 seats front runners elected 17.4 lowest percent of FP votes taken by succ. cand Kirkcaldy Central 3 seats front runners elected 19.1 lowest percent of FP votes taken by succ. cand Kirkcaldy East 3 seats front runners elected 16.3 lowest percent of FP votes taken by succ. cand Glenrothes West 4 seats front runners elected 19.1 lowest percent of FP votes taken by succ. cand Glenrothes North 4 seats front runners elected 11.7 lowest percent of FP votes taken by succ. cand Glenrothes Central 3 seats front runners elected 13.4 least percent of FP votes taken by succ. cand Howe of Fife 3 seats 14.1 lowest percent of FP votes taken by succ. cand

1 "turn-over" 17 percent FP-vote cand not elected, 14-percent FP-vote cand elected Tay 3 seats front runners elected 20.6 least percent of FP votes taken by succ. cand St. Andrews 3 seats front runners elected 11.8 least percent of FP votes taken by succ. cand. East Neuk 3 seats 15 lowest percent of FP votes taken by succ. cand

1 "turn-over" 19 percent FP-vote cand not elected, 15-percent FP-vote cand elected Cupar 3 seats front runners elected 18.9 least percent of FP votes taken by succ. cand Leven 4 seats front runners elected 12 least percent of FP votes taken by succ. cand Buckhaven 4 seats front runners elected 7.3 least percent of FP votes taken by succ. cand 1 "turn-over" 10.3 percent FP-vote cand not elected, 7-percent FP-vote cand (Labour) elected the turn-over Labour candidate likely received vote transfers from his running mate (elected) and from the Independent who came in first in the 1st Count with almost two quotas. =========================

effect of STV vote transfers one purpose of transfer is to allow a party to run as many candidates as it wants and suffer no pain (under usual circumstances - i.e. with agreement of voters) (without voter support, a party will not elect any, under any democratic system) transfers allow a party's vote to be concentrated on exactly the number of candidates that reflects the party's proportion of the vote. (with agreement of voters) (but the party must run that least the proportionate number of candidates (sometimes not all will be elected) and the appropriate number must survive to the point of election.) transfers also allow a voter's vote to be used to elect someone although not necessarily the voter's first choice. when the voter's first choice is not elect-able. in line with voters marked back-up preference, the vote may go to elect someone who does not belong to the party of the voter's first choice - if the first choice is not electable and if the voter marked preference for that "other party" candidate is the next usable back-up preference when the vote is to be transferred. transfers also allow a voter's vote to be used to elect someone although not necessarily the voter's first choice. when the voter's first choice is not elect-able. transfers also allow a voter's vote to be used to elect someone when the voter's first choice is elected. such a vote is declared to be surplus for that successful candidate and transferred on to where it has better chance to be needed. This allows intra party transfer so that a party's vote can be concentrated on just the number of seats it elects. it also allows broader extra-party co-efficiency - when one right-wing party no longer has candidates in the running (all have either been elected or eliminated) its remaining "loose" votes can move over to help candidates of another right-wing party and so on. with a party that starts out in 1st Count with two quotas and runs two candidates but only two candidates, the only way for the party to get two seats (its rightful due) is: - if both candidates are not eliminated at any stage of the counting, and - if voters for the party (making up the two quota) who gave 1st preference to the most-popular party candidate give back-up preference to the other party candidate. (this is not hard and fast rule -- if the party takes a bit extra than two quota and only a bit of the voters "wander off" with their back-up preference, then no harm is done. or a party with two quota (or even just less than two quota) may win two seats if it elects one with quota and the other has a plurality in the end of the count when quota is not required, or if both candidates hang on to the end and then win both seats without quota by having plurality if two seats are filled through plurality at the end (though that seldom happens - usually a party worth two seats takes one early and one later, and often only one seat is filled through plurality, at the end not two.). Another factor to consider is even if a party starts with less than two quota it may accumulate votes through transfers from other candidates (independent or belonging to other parties) enough to take two seat, either by taking quota twice or by taking a seat without quota but with plurality at the end. thus Gallagher index calculation that looks just at 1st Count does not see that a voter initially voting for A might due to the marked preferences, have the vote go to Party B, which might be just as much the preference as Party A. Certainly the way voter are pushed to vote strategically under FPTP is not considered when Gallagher is derived. ranked votes make the deal more open and thus because votes have more liberty to mark their vote for a party with less chance they do, thus all other things being equal, GI would be higher (worse) under ranked voting systems than under X voting. However, GI is not worse under STV - this is due to the better ability of votes even if truly placed, to be used effectively than under X voting system, where even if placed in such a way as to attempt to be used effectively, they are often wasted in great numbers. =============================== what percentage of votes is a winner under STV? Short answer from stats for Scotland's 2007 local government elections: successful candidates received range of percentage of votes in 1st Count - ranging from 50 percent (before surplus votes transferred away) to as low as 3 percent (before vote transfers received) so obviously below quota there is a grey zone where percentage may be enough to be elected but not necessarily. Only one was elected with as few as 3 percent - likely there were many, many who got more than 3 percent on the 1st Count who were not elected. ======================================== here's the stats for Scotland's 2007 local government elections (numbers are not exact) 190 wards had 3 seats quota is 25 percent (3-seat ward total seats 570) 163 wards electing four seats quota is 20 percent (4-seat ward total seats 652) total seats filled 1222 485 elected by first preference (by receiving quota on 1st Count) 737 elected after vote transfers, either by achieving quota or without ever attaining quota. 485 elected by first preference (in 1st Count) -- 215 in three-seat districts; 268 in four-seat districts 737 elected after transfers -- 351 in three-seat districts; 387 in four-seat districts percentage of votes in 1st Count - ranging from 50 percent (before surplus votes transferred away) to as low as 3 percent (before vote transfers received) so obviously below quota there is a grey zone where percentage may be enough to be elected but not necessarily. Only one was elected with as few as 3 percent - likely there were many, many who got more than 3 percent on the 1st Count who were not elected. (the percentage received by unsuccessful candidates is not known.) Many were elected who got between 15 percent and quota (25 or 20 percent in these wards) on the 1st Count. Some were elected who got between 10 percent and 15 percent on the 1st Count. Several were elected who got between 7 percent and 10 percent on the 1st Count. One was elected who got 5 percent on the 1st Count. One was elected who got 3 percent on the 1st Count.

source for this information is Electoral Reform Society of Scotland tables entitled "Elected Candidates' Shares of First Preference Votes in Ward" [ I think James Gilmour shared them with us maybe last year] other info on Scotland local authority STV is https://www.electoral-reform.org.uk/latest-news-and-research/publications/2007-scottish-local-elections/#sub-section-7 =========== A question that was raised recently was what quota meant under STV i mentioned that it is sometimes taken as effective electoral threshold any above are elected and any below are not. But in STV it works like this - any who get quota (or above) are assured of election; it is possible to be elected with less. mathematically it works that way whether you are talking about first preference votes or combination of first preference votes and votes that have been transferred from others. (votes landed on a candidate due to secondary preferences count for just as much as first preference votes - this is somewhat confused by fact that in some STV system votes transferred as surplus votes that had been received by a successful candidate are sometimes transferred as just fractions of themselves. but in all STV systems votes transferred from eliminated canaite are transferred as whole votes. and in Malta and Irish elections even surplus votes are transferred as whole votes (in the form of say 10 out of 100, while in those systems where fractions are used it is in the form of 1/10th of each of the 100). Time-wise the rule of "over quota elected, under quota sometimes yes sometimes no" is nuanced - any who get quota any time are declared elected. only at the end when last seats are being filled can someone be elected with less than quota. Due to "under quota sometimes yes sometimes no" rule, it was said that about 12 percent is enough to be elected in a five member district. in a five member district 17 percent is quota. to win with anything less than quota is iffy - in the first count, the top leaders are always elected either immediately by surpassing quota in first count or later by accumulating enough to pass quota, but the last say one two, or three of the five leaders in the first Count in a five-seat district may not be elected if they never get quota and if they are passed by other initially less popular candidates. (my analysis below is based on the 3-seat or four-seat districts used in Scotland's local authority elections in 2007. but likely 5-seat districts show at least as much range.) at the end when the field of candidate is thinned to the number of remaining one seats, plurality is enough to be elected. (quota no longer is relevant in some cases as the number of votes still in play has been reduced to where it is unattainable by any of the remaining candidates.) Plurality (relative popularity) is now the working mechanism (same as FPTP) but its application (held in contempt in FPTP) is acceptable in STV in the later stage for these reason- - sometimes only two candidates remain and one will have a majority of votes still in play and the other won't. at the end if one seat is to be filled, majority is all you can ask for. so many vorts have already been used to elect someone that the end process is not critical - in FPTP it is win at the beginning or nothing at all. - vote splitting has been resolved in most cases earlier - perhaps one candidate remains of each of the two largest parties (other(s) of the party have been elected, perhaps some eliminated) and perhaps a candidate of a third party that has somehow managed to stay out of the bottom so that he or she avoided being eliminated. for the largest parties it is not make or break -- each has won seat(s) already, the third party candidate is on life support - he or she has not won quota but if he or she has not stepped on too many toes (showing PR's much-vaunted polite elections), he or she may receive lower preferences from other small parties or even some supporters of candidates of large parties, to eke out a win getting one seat while the large parties take the lion's share of seats in the district.

So here's the main thing -- what percentage of votes is a winner under STV? (numbers are not exact) here's the stats for Scotland's 2007 local government elections 190 wards had 3 seats quota is 25 percent (3-seat ward total seats 570) 163 wards electing four seats quota is 20 percent (3-seat ward total seats 652 total seats filled 1222 485 elected by first preference 737 elected after vote transfers either with quota or without ever attaining quota. 485 elected by first preference (in 1st Count) -- 215 in three-seat districts; 268 in four-seat districts 737 elected after transfers -- 351 in three-seat districts; 387 in four-seat districts percentage of votes in 1st Count - ranging from 50 percent (before surplus votes transferred away) to as low as 3 percent (before vote transfers received) so obviously below quota there is a grey zone where percentage may be enough to be elected but not necessarily. only one was elected with as few as 3 percent - likely there were many, many with even more than 3 percent who were not elected. (this disparity is not as dangerous as under FPTP, because - surplus votes were transferred away from those most popular in the 1st Count. the success of some low-popularity candidates is due to intra-party transfers of votes. - every party likely gets benefit or suffers the same from the disparity, unlike FPTP where the leading party likely wins far more than its due share of seats while other parties get far fewer. and likely other reasons arising from the party proportionality produced by STV. peak votes (percentage of votes held when elected) -- unknown Scotland's three-member districts in 2007 local government elections 1st Count most-popular candidates 155 won on 1st Count with quota 1st ount 2nd-most-popular candidates 55 won on 1st Count with quota1st Count 3rd-most-popular candidates 5 won on 1st Count with quota 1st Count 4th-most-popular candidates 0 won on 1st Count with quota those who got quota after 1st Count 1st Count-most-popular candidates 35 won later (I assume by getting quota later) 1st-preference percentage for winners ranged from 50 percent to 20 percent 1st Count-2nd-most-popular candidates 130 won later (I assume by getting quota later) (five not win at all) 1st-preference percentage for winners ranged from 35 percent to 13 percent 1st Count-3rd-most-popular candidates 160 won later (I assume some by getting quota later; others never getting quota) (25 not win at all) 1st-preference percentage for winners ranged from 27 percent to 10 percent 1st Count-4th-most-popular candidates 26 won later (I assume some by getting quota later; others never getting quota) (164 not win at all) 1st-preference percentage for winners ranged from 20 percent to 7 percent Scotland's four-member districts in 2007 local government elections 163 4-seat districts quota was 20 percent

those who got quota in 1st Count 1st Count most-popular candidates 153 won on 1st Count with quota

1st-preference percentage for winners ranged from 50 percent to quota

1st Count 2nd-most-popular candidates 100 won on 1st Count with quota

1st-preference percentage for winners ranged from 34 percent to quota


1st Count 3rd-most-popular candidates 12 won on 1st Count with quota

1st-preference percentage for winners ranged from 23 percent to quota

1st Count 4th-most-popular candidates 1 won on 1st Count with just a little more than quota

Those who got quota after 1st Count 1st Count-most-popular candidates 10 won later (I assume by getting quota later) (all of them won ) 1st-preference percentage for winners listed here ranged from quota to 18 percent 1st Count-2nd-most-popular candidates 63 won later (I assume by getting quota later) (all of them won ) 1st-preference percentage for winners ranged from quota to 13 percent 1st Count-3rd-most-popular candidates 148 won later (I assume some by getting quota later; others never getting quota) (three of them not win) 1st-preference percentage for winners ranged from quota to 9 percent 1st Count-4th-most-popular candidates 146 won later (I assume some by getting quota later; others never getting quota) (44 of them not win) 1st-preference percentage for winners here ranged from quota to 7 percent 1st Count-5th-most-popular candidates 15 won later (I assume some by getting quota later; others never getting quota) (? of them not win - all wards had 5 candidates (or there would be no vote held) 1st-preference percentage for winners ranged from 17 percent to 8 percent 1st Count-6th-most-popular candidates 3 won later (I assume without ever getting quota) (? of them not win -- I don't know how many wards had 6 candidates 1st-preference percentage for winners ranged from 10 percent to 6 percent 1st Count-7th-most-popular candidates one won later (I assume without ever getting quota) (? of them not win -- I don't know how many wards had 7 candidates 1st-preference percentage for this winner was 5 percent 1st Count-8th-most-popular candidates one won later (I assume without ever getting quota) (? of them not win -- I don't know how many wards had 8 candidates First-Preference percentage for this winner was 3 percent

======================================= 2007 stats

353 separate STV-PR elections in the Scottish LA elections, in the 32 separate councils


I see in Aberdeen City, every ward, except one, elected one member in the first count.

the one ward that was exception had none win in the first stage.

so some wards  it seems, had two (or three?) win on the first stage. That is required if the 1+ average of candidates per ward elected by quota in First Count is to be produced.


To emphasize an old point, the chart could say 737 elected after transfers (versus "elected by transfer").

Like in the cases of the first-preference winners, many of those "elected by transfers" would have been elected even without the transfers due to being in winning position in the first count. Even if not achieving quota, they would have been elected by plurality at the end, if it had come to that, 


(In true fact, many of the 737 did not achieve quota even with transfers but were elected regardless --


in Aberdeen City, 11 of the elected 43 members did not achieve quota even after transfers.


of Aberdeen City's 43, only two were not in a winning position in the first stage.


general rule of thumb -- only ten percent of seats are filled by someone who come out of the lower ranks to take a seat.


that is not to say FPTP or BV would produce same fairness - 

Single voting in MMD is recipe for great fairness, and that is done in the first count, that transfers only polish or confirm.


(The A-Z "Supervote" circular does not not actually mention "single voting in MMD", as far as i recall. so despite its importance and almost self-sufficiency, it is an overlooked part of the STV process.)



2007 Scottish local authorities election Councillors elected by First Preference quota and after transfers

May 2007

Councillors elected

Elected by FP quota

Elected bytransfers

Percent elected by transfers

Conservative

143

58

85

59%

Labour

348

130

218

63%

Lib Dem

166

36

130

78%

SNP

363

201

162

45%

Others

202

60

142

70%

Scotland

1,222

485

737

60%

485 elected by First preferences (by quota in the First Count) is an average of about 1.5 per ward in the 353 wards.



That was for 2007.  The pattern may well be different for subsequent elections: 2012, 2017, 2022

===============================

4

1 view

Recent Posts

See All

Comments


bottom of page