Canadian use of STV and IRV partial and tentative but STV effective where used in Alberta and Manitoba, says Oxford Handbook of Electoral Systems

(work in progress)

The Oxford Handbook of Electoral Systems (2018)

CANADA (p. 741-762)

p. 748 says "BC is the only Canadian province where FPTP was ever discarded province-wide and replaced by another system."

he had previously mentioned that Alberta and Manitoba had adopted STV in Winn. in 1920 and then AV elsewhere in 1924.

And that Alberta had made that double change in one stroke.

so why is that ignored and BC given credit for doing it first, 27 years after Alberta?

p. 749 says UFA adoption of the STV/IRV mix was Machiavellian as it was strong where AV used and weak in the cities and thus needed the help that PR gave.

[But actually the UFA did not run in Calgary so got no benefit of STV in Calgary, and ran just one candidate in Edmonton and did win that seat, but likely would've anyway even under FPTP or IRV, as its candidate was the most-popular person in Edmonton, taking quota in the first round of counting, so not that weak after all. If the UFA had been greedy for seats, it could have run two candidates in each city and perhaps won two or three more seats than they did win.]

p. 748 draws four lessons on Canada's use of STV/IRV

1. no great upset when it was cancelled

2. use of PR in Canada confined to just two provinces and only urban areas in those provinces. never was more than 25 percent of members elected through STV. but it did show that where STV used, more proportional outcomes and it did allow smaller parties to gain representation.

3. [Regarding IRV] variant of IRV used was optional preferential voting (OPV).

   in Canada's use of STV, it used optional preferential voting (OPV) and that meant that the transfers had less effect than if compulsory preferential voting had been used.

4. [Regarding IRV] actual impact of lower preferences on electoral outcomes under IRV was minimal -- transfers mostly confirmed the victory of the candidate leading in the first count and therefore did not change anything, apart from delaying announcement of the final outcome. (The Oxford Handbook of Electoral Systems, p. 748)