Edmonton, the capital city of Alberta, used STV in city elections for a short time in the early 1900s. Despite the short-ness of its use, it proved itself able to produce mixed roughly proportional city councils.
Edmonton used STV from 1923 to 1927 - five elections. And in each election it produced mixed roughly proportional results that reflected most of the votes cast.
The city's elections were complicated by the fact that the southside was guaranteed two seats on council. Party labels were used - the business-oriented Citizens Committee and Labour Party - but STV elections do not use parties to determine successful candidates. Voters vote for individual candidates, and the vote tally of each candidate is compared to a "quota" and to the other candidates' tallies to determine the result. The election of multiple winners in a district ensure that most of the votes cast are represented by the successful candidates, if not through first preference then at least someone that the voter preferred over the others.
=====
After successful referendum vote in December 1922, Edmonton switched to Single Transferable Voting.
Very little had to be changed to bring in STV.
STV uses multi-member districts, and each voter casts just one vote, which is transferable - if necessary and if the voter has marked one or more back-up preferences.
In Edmonton's case, the electoral district did not have to be changed as Edmonton was already electing its councillors in one at-large city-wide district.
Previously, each voter had been able to cast as many votes as there were seats to fill, so a necessary change was to restrict each voter to just casting one vote. Votes had to be instructed in how to mark the preferential ballots. But many only marked their first preference - the only change for them then was to mark their choice with the number 1 instead of an X. Even without any back-up preferences, the voter was no worse off than under the Block Voting elections used in previous city elections or under First Past The Post single-winner elections used in federal elections at the time. If their one and only marked candidate was not elected, the vote was ignored same as it would be under those other systems.
But likely the voter, even if he or she marks just one candidate, is more likely to be satisfied with the result than under those systems.
Why?
Because under Block Voting or single-winner elections, one group can take all the seats,. Chances are that the voter - any voter - will not be among that one group so will not get no satisfaction from the result.
But under STV no one group can take all the seats so chances are that the voter - any voter - will find more satisfaction and see someone he or she votes for elected than under those systems.
Mayor and six aldermen were elected in the 1923 election. 13,000 votes were cast in this election.
In the aldermanic election, at least 8000 of these votes - about two-thirds of those who voted - were used to elect someone. (In the previous election, it is possible that no more than 5400 voters elected all the councillors. If that number of voters gave all of their votes to those who won, that number - less than half the voters - could have elected all the councillors that year.)
First preference votes cast for those who were successful in the end: James East (Labour) 3004 (had quota of 1693 in the First Count) Ambrose Bury 1993 (had quota of 1693 in the First Count) William Rea 1150 (eventually received quota 1693 votes) James M. Douglas (ss) 928 James Findlay (Labour) 707 (eventually received quota 1693 votes) Joseph J. Duggan (ss) 563 Douglas and Duggan were elected due to being the most popular and last remaining southside candidates. They had more votes than they started with but their final (peak) number of votes is unknown. Rea was last northsider elected so received the consolation prize election - to serve just a one-year term while the others were elected to two-year terms. Unsuccessful candidates and their first choice votes Thomas Ducey 789 Archie Rendall 702 J.J. McKenzie 420 Esther Saunders 380 (one of first women to run for city council) J. Boyd McBride 317 A.K. Putland (southside) 312 H.J. Pallot 308 Jan Lakeman 265 (perennial Communist Party candidate) total 3493 Most of these votes were transferred to other candidates, some of whom were elected. The successful candidates altogether had at least 8000 voters at the end. Four were elected with full quotas 4 X 1693 = 6772 Two southsiders were elected due to the southside guarantee. The southside candidates had at least 928 and 563 votes when they were elected. Time spent to do the transfers is often used as justification to retain the X voting systems but in this election it only took 13 hours for staff to do all transfers and announce the official results of the mayoral and aldermanic elections. And it was time well spent - it produced a much more balanced result than under Block Voting. The total number counted in the first count of the 1923 aldermanic election and then counted when they were transferred was less than 17,000, while in the 1922 election 53,000 votes were counted in the aldermanic election. So the work involved was actually less under STV than it had been under X voting in 1922. The number of spoiled or rejected ballots is also put forward as a major shortcoming of the STV system.
About 2500 votes were spoiled in the aldermanic election. The Dec. 12, 1923 Edmonton Bulletin said many of these ballots bore five or six number "1"s so it seems the change from the multiple voting of the old block voting system to the single voting of STV had not been adequately explained. Other of the "spoiled<' ballots were simply left blank. perhaps because the voter voted in the mayoral election but simply did not care who was elected for the aldermanic seats. The number of spoiled votes was higher than any previous election but its exact number is unclear. It seems though that their presence did not have an impact on the fairness of the election. The number of spoiled votes varied from the mayoral and the aldermanic contests. At least 1200 were spoiled in the aldermanic contest, where the number of candidates was larger than the mayoral contest. Some of the spoiled votes were ballots improperly filled out (say with an X instead of a number showing ranking), but some were declared spoiled for contests where the pertinent part of the ballot was simply left blank, such as no first-choice preference being marked for mayor or no first-choice preference was marked for alderman. The number of spoiled ballots was larger than any previous city election but the final result meant that only 3493 voters (out of 13,000) did not see their first choice elected. (Edmonton Bulletin, December 12, 1923, p. 1) Even as it was, more votes were used to elect someone than likely had been the case in the previous election. The Block Voting used in 1922 was not transparent and it cannot be known how many voters actually elected the six members. It could have been as few as 5400 or as many as all of the 11,000 who cast votes.
But under STV almost 9500 voters, out of the 13,000 voters who participated in the 1923 election, saw their choice elected, a very high rate of Effective Votes.
The same kind of high rate of Effective Votes was seen in provincial STV elections in Edmonton and Calgary starting in 1926, as discussed in other Montopedia blogs.
====================================
STV uses transferable preferential ballots. And some voters saw their first preference elected and some saw just their second choice elected, and some saw both elected. Others - less than 30 percent of the voters cast - elected none.
In 1923, 70 percent of the first-preference preferences marked on the ballots went to candidates who were elected in the end.
Candidates of two parties were elected. The business slate, the Citizens League, took the most seats but Labour did well also.
Within each slate the most popular individual candidates were elected.
Two candidates had quota on the first count and were immediately declared elected. (These were Labour's East and CC's Bury.)
Findlay achieved quota on the 2nd count, so that seat was filled early as well. One thousand votes that were marked for East as first preference and Findlay as second preference had both their first choice and their second preferences elected.
At the end of the 2nd Count, three empty seats remained. All five southside candidates were still standing and there were six northside candidates also still in running. At least two southside candidates would be elected to two-year terms. The counting of the vote would continue until: - the three seats were filled through three getting quota, two of them being southsiders, - there being only two southsiders and one northsider left standing (all of whom would be declared elected even without achieving quota), or - a northsider achieving quota, and there only being two southside candidates left standing to take the guaranteed seats. (The number of remaining northside candidates remaining would be immaterial as they could not be elected. However any votes that unsuccessful candidates possessed at the end would be ones that would not be used to elect anyone so would be counted as wasted.), or - a southsider achieving quota and sometime later there being just one northsider and one southsider left standing, or - a southsider achieving quota and eventually there being just two northsiders and one southsider left standing. The least-popular northsider would be eliminated, leaving the two remaining candidates to take seats. It took 10 counts for Rea, a northside candidate, to achieve quota. (Thus the third option listed above was the one that happened.) 10 counts may seem like a lot of work, but these counts involved relatively few votes. the 3rd Count saw transfer of only 265 votes, for example. Rea, one of the successful candidates, thus aggregated another 500 votes.
After Rea's election, there were only two seats left empty and these had to go to southside candidates.
Southsiders James Douglas (a former MP) and Joseph Duggan were declared elected to fill the two southside seats.
Rendall and Thomas Ducey (Edmonton baseball's man) had many votes in the first count but not quota and were not elected in the end. Five others, two from the southside, also were not elected.
The six successful candidates had received 8345 votes in the First Count so even without vote transfers the result showed a high proportion of the votes cast were used effectively to elect a councillor. East did not retain all of his votes. Some of East's votes went to Findlay; some went to other candidates who would eventually be declared elected; and some to those who would be eliminated. At the end, 8345 of the 12,000 voters who cast votes could look at the successful candidates and be happy about the result - the voter's first choice was elected.
Many voters were satisfied when their first choice, East, was elected.
A thousand voters could relish in the success of both East and Findlay. Findlay was the second choice of a thousand voters who had chosen East as the first choice. And then some of those votes were transferred on to Rea, Douglas and Duggan to help in the election of those. These voters - an unknown number - at this point in time were happy that their third choice was elected as well.
Of course each vote could only be used once. But the voter who cast his first choice for East, his second for Findlay, and later choices for Rea, Douglas and/or Duggan must have been well pleased with the overall election result.
Many others who voted for Bury as the first choice had a similar experience. If they had marked any of Findlay, Rea, Douglas and Duggan, and even East, as their secondary preferences, they could see those choices elected as well. But the vote itself could only be counted for one of them. If we add Findlay's vote total in the 2nd Count and Rea's 500-vote rise by the 10th Count (when he achieved quota), the number of votes received by the successful candidates exceeded 8000. four quotas 4 X 1692 = 6768 vote tallies of Douglas = 928* votes tallies of Duggan= 563* Total: 8259 * These totals are first-preference votes. By the time these two were elected in the 10th Count, Douglas and Duggan had aggregated even more votes. (But the number of votes they had cannot be known due to the official vote count chart being lost at this late year.)
So Edmonton's 1923 city election showed the fairness that STV could provide, that STV could ensure that only a small fraction of votes cast would be wasted, that votes cast by most of the voters would be used to elect someone.
The next four elections did the same fine job as well. STV would finally be discarded by voters but not because it was un-workable, nor because it was ineffective at producing balanced city councils that reflected votes cast by most city voters.
An upcoming blog will look at why in fact Edmonton's municipal STV was discarded.
=========================================================
eiSome sources say that spoiled ballots were a problem in Edmonton's city elections but the benefits of STV far out-weighed the problem of spoiled ballots.
The number of spoiled votes was higher than any previous election but its exact number is unclear. It seems though that their presence did not have an impact on the fairness of the election. The number of spoiled votes varied from the mayoral and the aldermanic contests.
At least 1200 were spoiled in the aldermanic contest, where the number of candidates was larger than the mayoral contest. Some of the spoiled votes were ballots improperly filled out (say with an X instead of a number showing ranking), but some were declared spoiled for contests where the pertinent part of the ballot was simply left blank, such as no first-choice preference being marked for mayor or no first-choice preference was marked for alderman.
The number of spoiled ballots in 1923 was larger than any previous city election but the final result meant that only 3493 voters (out of 13,000) did not see their first choice elected. (Edmonton Bulletin, Dec. 12, 1923, p. 1)
By comparison, in Edmonton's 2017 city election, held using First Past The Post, more than half the voters did not see their choice elected.
================================ see also Montopedia blog "Edmonton's STV elections 1923-1927"
====================================
Comentários