Much info on the 1919 Sligo election in this publication of the Proportional Represention Society of Canada.
The first municipal P.R. elections in the United Kingdom: Sligo (Ireland) municipal elections, January, 1919 : a practical demonstration of the working of the single transferable vote
published by the Proportional Representation Society of Canada, in [November] 1919.
This is said to be "P.R. pamphlet No. 8" [but I have not come across the previous ones in the series -- a search in CIHM for "Proportional Representation Society" and "P.R. pamphlet" yielded no others]
(CIHM No. 99422:
Says that at that point in time (November 1919), Calgary had used STV for two city elections and that STV had been used to fill the UK House of Commons University seats in 1919.
It says the January 1919 municipal election in Sligo, Ireland was the first use of STV in a municipal election in the UK.
And it prophesied that Sligo's successful use of STV would open the door to many more municipalities in UK using STV. Ireland soon got its independence from Britain, and many Irish cities as well as Republic of Ireland's Dail elections did begin to use STV.
But it would be many years before many UK municipalities adopted STV.
Scottish local authorities now use STV but that is a fairly recent development.
=======
from my "Timeline of electoral reform" blog:
(1919/1920 United Kingdom/Ireland -- 1919 Sligo city election used STV. first borough in UK to use STV.
Protestant-Catholic troubles, labour tension, borough in deep financial straits. Three 8-seat wards. Members of three parties (Ratepayers, Sinn Fein, Labour) plus Independents elected.
80 percent vote turnout.
newspaper hailed election as "absolutely fair."
Whole-vote "exact method" for transfer of surplus votes.
(Sligo election was so successful that STV was prescribed for all cities and towns in Ireland. On Jan 15, 1920, 126 bodies in Ireland elected by STV. (likely "Exact method" used to transfer surplus votes.) (Hoag and Hallet, PR (1926), p. 237)
===
*The PRS of Canada publication shows clearly (page 18) that Sligo used the "exact method" to transfer surplus votes.
West Ward
Sligo elected 8 in each ward in the 1919 election.
Thus in West Ward with 940 valid votes cast, the quota was 105 (1/9th plus 1)
Perry received 169 votes in the first count
His vote tally exceeded quota, and he was declared elected.
his surplus was 64
166 were transferable; three did not have next usable marked preference.
the "exact method" is used so transfer of the suplus is mathematical reduction of his transferable votes.
surplus/number of transferable papers = 66/166 = 32/83
of the 166, 77 were marked for Kerr
77 times 32/83 = 29 57/83 = 30 votes transferred to Kerr (57/83 being one of the largest fractions, the vote count was rounded up)
with 30 transferred to Kerr, 47 remained behind as part of Perry's quota.
the bundle of Perry's votes that were marked next for Kerr were brought up.
His 30 papers last filed, those on the top of the bundle, were transferred to Kerr and placed on the top of Kerr's bundle of original votes, which was already on the Returning Officer's table.
When the transfer of Perry's votes were conducted to the other candidates as appropriate to their portion of his transferable votes (all in all being 64 votes), Perry was left with 102 transferable votes, add in the three that were non-transferable and you have Perry left with 105, his quota.
Kerr previously had had 82 votes so the addition of the 30 transferred from Perry gave him more than quota.
his surplus was 7.
In the old-time application of the Exact Method, the procedure used to assess any later transfers, surpluses that arise when candidate exceed quota in rounds subsequent to the first Count, is different from those that arise from candidate exceeding quota in the First Count. While all the votes held by the First Count victors such as Perry were considered, for later winners (such as Kerr) only the "last parcel" received was used to assess the surplus transfer.
As the 1919 publication stated:
"In dealing with any secondary surplus, that is a surplus that arises from a transfer, only the papers which give rise to the surplus are examined for subsequent preferences. So in dealing with Kerr's surplus of 7 votes the 30 papers received from Perry were sorted according to the next choices thereon, in this case, the third choices; and the surplus was apportioned among the candidates entitled to benefit in accordance with the procedure already described."
Of Kerr's last parcel, of the 30 papers he received from Perry, the portions were:
1 to Hamilton (perhaps 4 of the last parcel)
1 to Minan (perhaps 4 of the last parcel)
5 to Tiche (perhaps 22 of the last parcel).
TOTAL: 7 (the number of surplus votes)
(Note that none of Kerr's first count votes were used to assemble the transfer, for which there are grounds for complaint against the "exact method." If all the votes held by later winners were used to assemble the surplus transfers (as was done for such as Perry) this cause for complaint would be removed.)
No candidate received enough of Kerr's transfers to exceed quota.
so vote count proceeded along different lines thereafter:
"At the conclusion of the third count two Cndidates had been elected, and there were fourteen candidates in the running for the other six places, none of whom had obtained a quota. The Returning Officer accordingly declared defeated the candidate lowest on the poll, Thompson, with 16 votes. His papers were sorted according to the next available preferences.
The term "next available preference" needs definition. In nearly every case the second choice marked on Thompson's papers was the next available preference; but, where the elector had marked either of ths elected candidates. Perry or Kerr, as second choice, the third choice became the next available preference; and when an elector had marked Perry or Kerr as his second choice and Kerr or Perry as his third choice, the fourth choice became the next available preference."
Finally after 12 Counts, the eight seats were all filled.
Party proportionality
judging by 1st Count preferences alone we see good party proportionality of the three parties involved in this race. (No Independent candidates ran in this ward this time.)
As the PRS of Canada publication stated:
"The Ratepayers polled four quotas and 86 votes over; they obtained four seats,
Sinn Fein polled two quotas and 23 votes over; they obtained two seats.
Labour polled one quota and 96 votes over (nearly two quotas), and obtained two seats.
The four Ratepayers' candidates and the two Labour candidates elected were those with the most first preferences in these two groups.
In the case of Sinn Fein, Flanagan, who had 44 first preferences, was elected, whereas Hughes (also of Sinn Fein), with 56 first preferences, was defeated.
Furthermore, Flanagan (Sinn Fein), who was elected, had fewer first preferences than Hamilton (Ratepayers), who was eliminated."
But at time of Hughes' elimination, Flanagan had accumulated more preferences than him.
The explanation for Flanagan's success comes from the transfers.
When Roche and Hughes were eliminated, more of their votes went to Flanagan than went to Hamilton. This makes sense as Flanagan, Roche and Hughes were all of the Sinn Fein, while Hamilton was not.
When Hughes was eliminated, already Flanagan had accumulated a higher voter count, and that is why Hughes, not Flanagan, was eliminated. Flanagan received many vote transfers from Roche, just before the elimination rota arrived at the lower of Hughes and Flanagan, and that put Flanagan a bit ahead of Hughes, who had to be eliminated. Saved from the 11th Count elimination, Flanagan then received many of Hughes' vote transfers and that put him over the quota. It also put him past Feeney who had been leading ahead of him up to that point. Feeney, as the last surviving SF candidate, hung on to win in the end as well. giving the Sinn Fein two seats.
At the end, Hamilton (of the Ratepayers) was the last unsuccessful candidate and his votes were not transferred. You would think that that might have meant the the Ratepayers would get one less seat than their vote share, due to this waste. But the Ratepayers had 86 votes over and above the four quotas and so the waste of 89 votes due to Hamilton's position did not make them receive less than the four seats.
Labour on the other hand was under two quotas but were closer to two quotas than Ratepayers was to five quotas. So in all fairness Labour got two seats and Ratepayers "only" four.
Every voter who voted in the West Ward saw at least one member of the party of their choice elected in the West Ward.
Among Ratepayers 420 first preference votes went to those who won in the end.
Among Sinn Fein 99 first preference votes went to those who won in the end.
Among Labour 150 first preference votes went to those who won in the end.
These votes were never transferred, except in so far as the candidate's surplus votes were (in all cases except in the case of Hande and Feeney).
As well, Hamilton's 51 first-preference votes were never transferred.
Some votes may have been transferred as many as six or more times, while the bulk of the winners' votes were not ever transferred.
Thus, 670 of the 940 votes saw their first choice elected and all saw at least one other of the same party elected as well.
At the end, about 800 votes (of the 940 votes cast in all) were used to elect someone, a much better rate than the rate of 50 percent or so seen in FPTP elections in general.
Although 12 Counts were done before all the seats were filled, Perry knew when first results were announced that he had won as he had quota. With quota, there was nothing any other candidate could do to stop him - he was immediately declared elected.
As well, most or all of the leading eight candidates knew they would likely go on to be elected in the end.
Those with less than half quota in the 1st Count likely knew they had no chance to be elected.
The transfers were useful in several ways -
they allowed a party's votes to be concentrated among the right number of candidates, if votes marked their back-up preferences that way.
This was necessary if one candidate was very popular and had surplus votes that would otherwise be wasted,
and was also necessary if a party put more candidates in the field than its vote share, so as to save the party from the bad effects of vote splitting.
But the transfers had little impact on the candidates in winning positions in the first count. (More on this below.)
Cross-voting -- transfers of votes across party lines
Perry's transfers showed a fair amount of cross voting.
This perhaps was due to all the other candidates still standing and being able to receive transfers.
Or perhaps he was a front runner and many voters wanted to vote for the front runner but his popularity was not based on party allegiance.
Effect of transfers
We see the 1st Count rankings and the final result:
1st Count votes final vote counts
Perry (RPRS) 169 105 elected (1)
Depew (Labour) 94 106* elected (4)
Connolly (RPRS) 91 106* elected (5)
Kerr (RPRS) 82 105 elected (2)
Tighe (RPRS) 79 105 elected (3)
Hande (Labour) 56 95 elected (7)
Hughes (SF) 56 (73 at peak) 0 at end
Feeney (SF) 55 91 elected (8)
= (above were in winning positions in the 1st Count) =
Hamilton (RPRS) 51 89
Flanagan (SF) 44 105 elected (6)
followed by about six more less-popular candidates
abbreviations: RPRS Ratepayers (businesss group) SF Sinn Fein
*Depew's and Connolly's surplus was never transferred as the one-vote amount never would've made a difference to the ranking of the other candidates. Or possibly all of their votes were non-transferable.
Thus we see that without the transfers, Hughes would've been elected, and not Flanagan. So only one initially-lower-ranking candidate rose up and passed other candidates to take a seat due to transfers.
Other than that one case, all the seats were filled by candidates who were in winning positions already in the 1st Count. (This is not unusual - in every STV election all or most of the candidates in winning positions in the 1st Count are elected in the end. It is calculated that only one in ten on the average are switched out.)
The PRS of Canada publication had this to say on the matter of the effect of transfers:
"In view of the comparatively small change in the relative position of candidates of the first and final counts, the provision for transferring votes may be regarded as of little value. But if the single vote were not transferable, the whole character of an election would be changed.
In the Sligo elections, every citizen felt that he was free to express his preferences, many or few, just as he desired. Every party, too, was free to nominate as many candidates as it wished. If the single vote had not been transferable, all this freedom and elasticity would have disappeared. In order to secure representation, minorities would have been compelled, as under the system of the cumulative vote, to calculate carefully their probable strength, to reduce the number of their candidates accordingly, and to issue precise instructions as to the candidate for whom different sections of their supporters should record their single vote. The election would have been rigid in all its aspects: it would have presented a complete contrast to the freedom and elasticity associated with the transferable vote."
So even though vote transfers make little effect, by doing the transfers the accuracy of the 1st Count rankings are tested and found to be true, and if not found to be true, are changed so a more-generally-popular candidate is elected over the relatively-unpopular 1st-Count leader.
===========
Public information
This was Sligo votes first experience at voting with ranked votes.
John H. Humphreys gave a well-attended instructional on the topic.
At it, interested persons expressed many concerns.
The PRS of Canada publication said:
"A question which in different forms was repeated several times was: "Does it pay to plump?"
In other words, would the elector gain by putting the figure 1 against his first choice and leaving the other candidates alone?
The answer was: 'No, the elector would not gain, he might lose.' His first choice had first claim upon the vote in any case. It would he transferred if his first choice had too many votes, or if his first choice was at the bottom of the poll and was declared defeated. If the voter did not express a second choice, the vote could not be transferred and might be wasted."
Anoherquestion asked was how many preferences sjhould be marked?
The answer was equivocal - the voter should rank the candidates down to where he or she has no preference about who is elected.
but a more specific answer was with eight to be elected, it would be wise to mark as many as 12.
===========================
Report on the 1919 election --
"Fair representation and no difficulties"
J. McCarthy, Returning Officer at the election of the Sligo Municipal Council by the single transferable vote, held in January, 1919, states:
The experiment of applying the single transferable vote to the municipal elections in Sligo has shown--
1. voters had no difficulty in grasping what they had to do.
2. In the past, owing to lack of interest on the part of a number of the electors, the polls were small. In the STV election, nearly 80 percent of eligible voters voted.
3. The count is not perceptibly longer than the old method in point of time where 16 candidates were competing for 8 seats.
4, at the count neither the candidates nor their agents had any trouble m following the various steps although It was their first experience of the system. The difficulty of the count appears much more formidable on paper than in actual practice. A systematic method eliminates the alleged difficulties.
5 So far as these elections are concerned, the results of the system have succeeded in making good the claims of the advocates of the system.. The rights of majorities are preserved and at the same time representation given to minorities. The various parties got representaton in a proportion wonderfully near to their voting strength.
6 The system is a practical one and the results have been received with general satisfaction. (page 18)
=======================
in a press comment in the Sligo Champion, reporting on the first use of STV in Ireland in January 1919:
"The system has justified its adoption. We saw it work; we saw its simplicity; we saw its
unerring honesty to the voter all through; we saw the result in the final count; and we join in the general expression of those who followed it with an intelligent interest - it is as easy as the old way; it is a big improvement and it is absolutely fair."
quoted in the "The 2024 case for the Supervote" essay by David Green.
======================================================================
Kommentare