The first count of votes in a STV election pretty much determines who will be elected in the end. But that result is not certain. Transfers caused by eliminations of least-popular candidates may create change. If a first-count leader is not as widely acceptable to voters as another candidate, he or she will not be elected. The transfers are more safeguards of democracy than anything else.
Looking at the STV elections of MLAs in Edmonton and Calgary from 1926 to 1955, we see this to be true.
The first count in STV elections - before any transfers were made - always produced a crop of leaders that was mixed, belonging to three or more parties.
One or two in each election already had quota on the first count and were declared elected.
The rest of the first-count leaders were basically the same as the candidates who were elected after the subsequent vote transfers caused by the elimination of the least-popular candidates.
In a couple elections no change was made through vote transfers. The leaders in the first count all were elected in the end.
In most elections, one or two first-count leaders were ultimately not elected - others less-initially-popular rose up due to wider acceptability to pass the narrowly-popular front-runners.
On two different occasions (the 1935 and 1955 Edmonton elections), three of the first-count leaders were not elected, but these were unusual cases.
Some of these changes resulted from the election of candidates of new parties not previously represented among the first-count leaders. But many of the changes were the replacement of a candidate by a different candidate of the same party.
The election of a candidate of a party not previously present among the leaders in the first count occurred in seven of the eight STV elections in Edmonton and six of the eight STV elections in Calgary.
The occasions where formerly lower-ranking candidates were elected:
In 1926 Labour Fred White rose from seventh place to take one of Calgary's five seats. In the first count White was the second-most popular of Labour's three candidates.
No Labour candidate was among the leaders in the first count.
In 1935 Conservative John Irwin rose from eighth place to take one of Calgary's six seats. In the first count Irwin was the most popular of Conservative's three candidates.
In 1935 Edmonton: there were 4 SC-ers, a Liberal, and a UFA candidate in the leaders in the first count. No UFA was among the six final winners. A Conservative was elected through vote transfers.
Three first-count leaders - SC Hall and Kennedy, and UFA's Lymburn - were not among the candidates elected.
Conservative D.M. Duggan rose from seventh place to take one of Edmonton's six seats. In the first count Duggan was the most popular of Conservative's five candidates. Two Liberal candidates also rose up to take seats. Van Allen had been eighth in the first count; O'Connor had been 12th in the first count.
Liberal Howson won a seat on the first count. Van Allen, the second most-popular Liberal in the first count, received many more of Howson's surplus votes than anyone else. He did not have wide appeal and, compared to O'Connor, did not receive many votes from transfers later.
O'Connor, the fourth most-popular Liberal candidate in the first count, received some votes from Howson, but more from the elimination of Liberal Conroy and UFA-er Lymburn.
Conroy had been more popular than O'Connor until halfway through the process when O'Connor accumulated more votes than him - more of those who voted for Liberal candidate Morrish had marked their next choice for O'Connor than for Conroy. Thus it became Conroy and not O'Connor who was eliminated when that point came.
(Interestingly, eventually the field of candidates thinned down to just four, all fairly close in vote totals. At that point there were only three seats open (Liberals Howson and Van Allen and SC Barnes having been elected), so the elimination of the least-popular candidate determined the success of the remaining three. The lowest-place candidate, SC Hall, was eliminated, and Conservative Duggan, SC Mullen and Liberal O'Connor were declared elected in this fashion. Hall had been a leader in the first count but had not received as many votes through transfers as O'Connor or Van Allen, who had not been leaders.
Duggan and O'Connor had not been among the leaders in the first count but had accumulated enough votes to avoid elimination and thus survive until the final designation of seats.
O'Connor received most of his added votes through transfers following the eliminations of Liberals Marshall and Conroy, and UFA Lymburn. (Votes transferred after the elimination of Conservative, Labor, Communist and Social Credit candidates did him little good.)
Duggan, the most popular Conservative in the first count, mostly accumulated votes from the elimination of lower-ranking Conservative candidates and from the elimination of Lymburn. These were enough to secure his survival to the end when he won a seat, to be the sole representative of the Conservative cause among the six city MLAs.)
In 1948 there was no Liberal among the leaders in the first count in Edmonton. The Labour man near the top of the count -- Morrison -- did not win a seat. Instead, Liberal Hugh John MacDonald rose up from seventh place to take one of Calgary's five seats. In the first count Hugh John MacDonald was the most popular of the five Liberal candidates. By 1948, most Labourites were running under the CCF banner. Morrison's candidacy under the Labour banner was unusual - no one else ran under the Labour label in this election - and not a winning strategy apparently.
In 1952 Calgary: there were no Liberals among the leaders in the first count. Independent Labour candidate D.F. McIntosh and Social Credit-er Glen did not win seats even though they were among the leaders in the first count. A different Social Credit candidate and a Liberal rose up to take their seats. McIntosh was the only Independent Labour candidate so that brand of sentiment was not represented due to the change made by the vote transfers.
In 1955 Calgary: there were 3 SC-ers, 2 Conservatives and a Liberal among the leaders in the first count. All these parties were represented in the final winners, but first-count leader Conservative Brecken was not among them. Liberal candidate Grant MacEwan, eighth place in the first count, rose up to take his seat.
--------------------------------------------
Vote transfers sometimes reduced the range of party representation in these elections compared to the first count leaders
This would seem to be a lessening of proportionality but that is not the case. In all cases the second or third candidate of one party that won had more support than the candidate who by losing his leading position, lost his party any direct representation in the legislature.
STV does not create representation by electing the widest possible range of candidates, giving each party one seat, but instead by electing them proportionally. A party that has twice the votes of another is given two seats compared to the other. Proportionality means each successful candidate should be elected by about the same number of votes.
In 1940 Calgary: Fred White, now a CCF candidate, was denied a Calgary seat even though he had been among the leaders in the first count. He was the only CCF-er among the leaders in the first count. A Social Credit candidate rose up to take his seat.
In 1940 Edmonton: J. Percy Page, a candidate for the Independent Citizen's Association, and Liberal Lazarowich did not win Edmonton seats even though they were among the leaders in the first count. Two Social Credit candidates rose up to take their seats. Liberal J. Harper Prowse was also among the first count leaders and went on to win a seat, so the Liberal party was still represented in the final result. Page's ICA was excluded there. Page was the only candidate for his Conservative-oriented group in this election. )
In 1955 Edmonton: there were 4 Liberals, a SC-er, a CCF-er and a Conservative in the leaders in the first count. All these parties except the CCF were represented in the seven final winners. Three first-count leaders were not among the candidates elected - Liberals Dechene and Bailey, and CCF-er Roper. A different Liberal (Tanner) rose up to take a seat. Two SC candidates also rose up to take a seat, in part due to vote transfers from the SC leader in the first count - Ernest Manning who alone received more than a quarter of all the votes cast in the city in the first count, far more than he needed to be elected.
--------------------------------------------------
An example of an election deeply affected by vote transfers but with no effect on party representation (final results versus first count)
STV election in 1952 in Edmonton. Seven MLAs elected.
The leaders in the first count were 3 SC-ers, 2 Liberals, a CCF-er and a Conservative. All these parties were represented in the final winners, but two specific candidates among the first-count leaders were not among them - Liberal Dechene and a Social Credit-er Holowach. A different Social Credit candidate and a different Liberal rose up to take their seats.
-------------------------------------------------
Scottish Local Authority elections show the same.
In 2007, there were 190 three-member wards and 163 four-member wards in the municipalities, .
In the 190 three-member wards across Scotland's municipalities, only 30 candidates who placed fourth or fifth in the first count improved their position enough to be elected. Thus, of the 570 front runners in the first count in the 190 three-member wards, only 30 of them (five percent) failed of election.
In the 163 four-member wards, only 20 who placed fifth or sixth in the first count improved their position enough to be elected. Thus of the 652 front runners in the first count in the 163 four-member wards, only 20 of them (three percent) failed of election.
That was about the same as when STV was used in Canada
So the first count in an STV election basically sets up which parties will be represented, and which candidates will win in the end. Vote transfers may make some differences, say, if a party's votes were enough for a quota but were spread over many candidates, each with not enough to win or if a party's votes were mostly at first given to one candidate, or if the supporters of two parties share their back-up preferences to allow one of their collective number to be elected.
If nothing else, vote transfers prove or ensure that the final winners have the support of most of the voters (75 to 90 percent of them). This is much fairer representation than FPTP where usually more voters are unrepresented by the winning candidates than are represented by them.
Thanks for reading.
==========================================
Comments