This page has 2 parts. The first is a local nearly-verbatim copy of the Globe & Mail news story by Gloria Galloway. The second part is me (Warren D. Smith) going through the Abacus/Broadbent report that was the primary source for that news story, saying some things I think about it (part summary, part reaction). Other facts about what people think about voting systems: 1, 2, 3.
By GLORIA GALLOWAY, The Globe and Mail, Ottawa 2 Dec 2015. Original story (includes online comments).
Most Canadians believe the federal electoral system needs to be changed, and those who advocate reform want to ensure that winning parties occupy the amount of space in Parliament that is proportionate to their level of support across the country, a new poll suggests.
With the Liberal government of Justin Trudeau promising that the election of Oct. 19 will be the last conducted under the first-past-the-post system, the Broadbent Institute has commissioned a study to assess what Canadians think about the way they vote. The survey by Abacus Data is to be released Wednesday morning but an advance copy was obtained by The Globe and Mail.
The polling firm surveyed 2,986 adults, randomly selected from a large representative panel of Canadians, to find out how they believe elections should be restructured, and also to estimate how the results of the most recent vote would have differed under proportional representation and under a ranked ballot. Those are the two systems that Mr. Trudeau has said he will ask an all-party Commons committee to study as the government decides, within the next 18 months, which way to go on electoral reform.
A survey of the size conducted by Abacus is expected to reflect the broad opinions of the Canadian public within 1.8 percentage points 19 times out of 20.
The poll suggests that a ranked ballot would have given the Liberals – who won a majority despite receiving only 40 per cent of the popular vote – an even greater number of seats in Parliament. But proportional representation would have reduced them to a minority and given Mr. Trudeau about 40 per cent of the seats.
Under a ranked ballot, the Liberal seat count would have climbed from 184 to 217, the Conservatives would have dropped from 99 to 66 and the NDP would have inched up from 44 to 50, the poll suggests. But, under proportional representation, the Liberals would have fallen to 136 seats, the Conservatives would have climbed to 108 and the NDP would have jumped to 67.
When asked to rank the top goals of a voting system, the respondents at large said they wanted simple ballots, strong stable governments, the ability to directly elect the MPs who represent their constituency, and assurances that the government has MPs from each region of the country.
But among those who said they want change, the top goal was a system that ensures that the number of seats held by a party in Parliament matches its actual level of support across the country.
One of the overriding messages of the survey is that Canadians are not satisfied with the current system, said Rick Smith, the executive director of the Broadbent Institute, a left-wing think tank that was founded by former NDP Leader Ed Broadbent.
"About 40 per cent of people said yes, they would be up for major reforms. About 40 per cent of people said they would be up for some reforms. And the rest of folks weren't up for much," said Mr. Smith. "So that's a pretty healthy constituency of people looking for change."
Those surveyed were asked to rank four different electoral systems:
Overall, 44 per cent of the respondents picked one of the two systems of proportional representation, while 43 said they were happy with first-past-the-post, and just 14 per cent said they liked the idea of a ranked ballot.
"There is some understanding that, if you have a stand-alone ranked ballot system, that alone will not give you a proportionate outcome," said Mr. Smith.
But the fact that 43 per cent of the respondents said they are happy with the way they currently vote, at the same time that 80 per cent said they are looking for some change, suggests Canadians are unsure about what should happen with the electoral system – and that Mr. Trudeau has his work cut out for him as he embarks on a process of reform.
This poll was badly written – and it also was conducted online, which is a bad way to conduct polls, although they claim their sample is representative anyhow. The result is unnecessary confusion and suspicion about what it tells us. But for what it is worth here it is. 2986 Canadians aged≥18 surveyed 3-6 Nov 2015.
1. The system Canada uses to elect MPs needs only minor changes (41%), needs major changes (33%), no change needed (17%), total revamp (9%).
2. Canadians care about electoral reform how much? 12, 36, 40, 12% for four possible caring levels from most to least (worded in an incredibly asinine manner).
3. Choose 5 of the most important goals for the new voting system from a list of 15 pre-canned goals. Ranked in order from most-checked to least – note with random 5-checking pollees each of the 15 would have been expected to be 33% checked, so above 33% is wanted more than average, and below 33% is wanted less (I drew a line at 33%):
(Total=500%.)
Goal e was presented as 2 slightly differently worded goals, which was idiotic and may have produced a "vote split" rendering this count of 38-41% perhaps merely an underestimate. The truth might be up to 65%. Or might really be only 38%. (Also e & g and i & b are somewhat clones also.)
Goal j was asinine. I also do not really know what h meant.
Goals a↔e and b↔e and i↔e and j↔e conflict, which the pollees may not have realized.
Goal e implies goal g. Goal i implies goal b, one would think.
Goal d is interesting and neither PR nor single-winner schemes that I know of, will accomplish it. A new kind of system would be needed to accomplish this.
Goal f also interesting and is not really about an election system, it is more about an un-election system.
4. Here are 4 system-types:
The pollees were asked to rank-order these 4 choices.
| System | #top rating it | 2nd | 3rd | #bottom rating it |
|---|---|---|---|---|
| MMP | 38 | 26 | 20 | 16% |
| LPR | 23 | 27 | 27 | 23 |
| SMP | 19 | 25 | 30 | 25 |
| IRV | 20 | 21 | 23 | 30 |
The idiots in charge of this survey failed to give a pairwise table, and failed to ask for ratings, only rankings. Nevertheless it is fairly clear MMP is the winner among those 4 choices, and would win with score voting if pollees gave their Kth choice score SK for any four fixed score values S1>S2>S3>S4≥0. Also, clearly LPR is the 2nd-most-wanted, i.e. it would have come second with score voting using any 4 fixed scores. Finally, IRV versus SMP is unclear – they are in the last two places but which one is bottommost is not clear. (Probably IRV is bottommost, though, since a ±1 change would suffice to make SMP beat it in the same any-4-scores sense.)
Presumably, then, any clear improvement over the MMP scheme described by Abacus would be even more popular.
5. "I always vote even when I believe my vote will not change the outcome" strongly agree=51%.
6. I have voted (dishonestly) for a party that was not my favorite. Strongly agree=19%, agree=27% (total 46%). Disagree=27%, strongly disagree=26% (total disagree=53%).
7. Abacus/Broadbent then tried to estimate what would have happened in 2015 election if those other voting systems had been used. IRV would have "produced an even larger false majority" for the Liberals.
With PR, the liberals would have a seat plurality but fallen 31-34 seats short of a majority, so there would have been a coalition.