It looks like we’re leaning towards a minority or coalition government. Under the circumstances, that may be a positive thing.
Hey folks, sorry for my delay in responding. My job had sucked all the time and energy out of my thought process so I finally had some time, between turkey bastincs, to sit down and smash out a bit of code to generate some actual, meaningful results for a more proportional voting system.
If reading long things isn’t your jam, then all you need to know is that my system needs more work, but I’m done poking at it for now so you don’t need to block me for being a bore (yet). Otherwise, let me prattle on for a bit after the cut.
To recap:
–A party provides candidates in each riding, as before.
–The voter votes for the candidate in their riding, as before.
– Seats are allocated based on a fraction of the popular vote: votes/total_votes * ridings (338) = total seats, rounded off.
–Seat allocation is that party’s top X showings, by percent. (ie: If party X has 10 seats, then their ten MPs are selected from their top ten ridings based on overall vote percent in those ridings.
Using 2015’s data:
Lib: 133.4 (133) down 51 from 184
Con: 107.9 (108) up 9 from 99
NDP: 66.7 (67) up 23 from 44
BQ: 15.8 (16) up 6 from 10
GRN: 11.6 (12) up 11 from 1.
Total: 336 seats (vs. 338)
Seat threshold (ie: what % of the riding’s vote did a party need to secure a seat)
Lib: 45.1%.
Con: 39.5%
NDP: 30.2%
BQ: 27.6%
Green: 8.9% (NB: In one riding, the Liberals had won by over 54%!)
Ridings with two MPs: 43
Ridings with no MPs: 45
Other trivia: The Libertarian party managed 0.7 of a seat, over 72 ridings. If they fielded the full 338, odds were decent they would have had at least one seat.
This system of mine seems relatively generous to the Greens, who only secured over 30% of the vote in two ridings. I’m sure they wouldn’t complain too much.
The spreadsheet summarizing the details can be found at ao.kalimonster.net/results2015.xlsx, if you care for that sort of thing. If you don’t care for that sort of thing, it’s still there.
I do kind of admit I was hoping there’d be relatively few negative attack ads, but I’ll own it, I’m a hopeless idealist on occasion.
I’m totally not surprised if the Conservatives feed Scheer to the leadership wolves if he doesn’t lead. It’s kind of standard operating procedure for both the Liberals and the Conservatives. Either way, I’m kind of glad to see we’re on the homestretch. It’s felt like a long campaign. I know, compared to the election campaigns south of the 49th this is just a blip, but still!
How does this account for over representation fo regional parties like the BQ? Since they are only running for ~ 25% of the available seats, does that get factored in somehow?
With respect, I’ll have to ask you to define “over representation” quite precisely.
Any party, regardless of whether or not I like their politics, would earn votes/totalvotes*338 seats, exactly like any other party. The seats they get is a direct reflection of the votes they earn. If any particular party has a strong showing at the election, it’s because they’ve spoken to a perceived need of the electorate.
Let me ask it this way.
Suppose two parties each get 25% of the vote nationally.
One party has only campaigned in one province and racked up huge majorities in that province, to earn 25% of the national vote.
The other party didn’t get any votes in that province, and its votes are distribuée pretty much evenly across the other nine provinces, averaging about 2.7% in each of the other nine provinces.
Both parties got 25% of the vote nationally, but one of them never got more than 2.7% in any province.Do they each get a quarter of the seats in the Commons in your model?
Northern Piper explained it.
The problem I have with various voting regimes and moving to a more coalition driven government is the problem of accountability for policy. For myself a key piece of democracy is direct attribution for government policy to a party and the ability of the population to have it replaced peacefully. Coalition governments explicitly distribute blame/praise thereby diluting the accountability I want to see.
Unless I’m missed my something, your model doesn’t take federalism into account, and allocates seats entirely based on the national popular vote.
But each province is constitutionally entitled to a set number of seats. What if the “Centralism Party” racks up huge vote share in Ontario, so much so that to get their equivalent proportion of seats, their votes have to be allocated to claim seats from other provinces for the “Centralism Party” even though other provinces have voted for the “Regional Party”? Would that allow the voters in Ontario to override the voters in smaller provinces, contrary to their seat guarantee?
The problem with this system, as Piper has pointed out, is that it is unconstitutional as you have described it. Every province has to allocated a certain number of seats. HOW you elect them is not Constitutionally prescribed, but the fact you have to elect them from particular provinces is.
If you go to simple proportional representation, which you have, you must specifically determine which 14 MPs are from Saskatchewan, which 4 of those people are from PEI, which 11 are from Nova Scotia, and so on. So are you doing it proportionally to that province’s vote?
If you do it proportional to provincial vote then regional parties become immensely more powerful and more attractive to create. You also get into some pretty hacky math with the smaller provinces; Ontario, with a minimum of 95 seats (it currently has 121, but 95 is the constitutional bottom limit) can allocate MPs pretty fairly - if you get 10% of the vote you’ll currently get 12 seats - but in New Brunswick, the fractions are a bit clunky.
The problem with this system, as Piper has pointed out, is that it is unconstitutional as you have described it. Every province has to allocated a certain number of seats. HOW you elect them is not Constitutionally prescribed, but the fact you have to elect them from particular provinces is.
If you go to simple proportional representation, which you have, you must specifically determine which 14 MPs are from Saskatchewan, which 4 of those people are from PEI, which 11 are from Nova Scotia, and so on. So are you doing it proportionally to that province’s vote?
You don’t have to. You could allocate MPs based on 338 seats and then just pick 78 people from Quebec, 34 from Alberta, and so on. But how do you decide which 34 represent Alberta? It would be bizarre if Albertans voted 50% Conservative, but you decided to appoint all the NDP MPs from Alberta but none from Ontario. So you could appoint the MPs based on the vote total in that province but mathematically that is not far from holding 13 separate proportional votess.
… but if you do it proportional to provincial vote then regional parties become immensely more powerful and more attractive to create. You also get into some pretty hacky math with the smaller provinces; Ontario, with a minimum of 95 seats (it currently has 121, but 95 is the constitutional bottom limit) can allocate MPs pretty fairly - if you get 10% of the vote you’ll currently get 12 seats - but in New Brunswick, the fractions are a bit clunky, and of course the territories only have one seat each.
I would also point out that under a pure proportional vote system you are giving up the idea of local representation in any sense that matters. It is no longer assured I will have an MP local to me, or that the people of Trois-Rivieres, Calgary Confederation, or Kingston and the Islands will have one. I assure you that absolutely does matter; my sister works in a constituency office, and local representation is important.
Those problems are easily be mitigated though. MMP (Mixed-member proportional representation) was designed so to keep local representatives and always ensure a representative parliament.
You get one local vote (first-past-the-post) and another party vote (pure proportional representation). The local vote gives you 50% of your parliament and the party vote offsets the FPTP to ensure a representative parliament.
Of course there’s other PR systems which can ensure representation (while keeping local representatives) too.
BC is turning out to be a 3-way battleground. It’s amazing that the Greens are still competitive in such a tight political environment.
Let me start off by saying that I’m not really married to this system I talked about, but I’m happy enough to talk about it. I’d be interested in solutions to problems we reveal.
If I understood you correctly, your scenario is this: The Central Party has swept Ontario, scoring effectively 25% of the vote nationally. The Regional Party has also scored 25% of the national vote, but distributed across the rest of the ridings.
So to make this happen, the Central Party scores an average 64% vote share per riding in Ontario, and 0% elsewhere, coincidentally engineered to get a 25% federal vote share. (64%Ontario_vote_fraction39%Ontario_population_fraction338 ridings) This winds up being ~84 seats. Ontario has ~120 seats available, which means all the MPs the Central Party has elected will sit in Ontario.
Since the Regional Party has 25% of the federal vote, they’ll get 25% of the seats as well, or ~84. These votes are distributed evenly across all non-Ontario ridings. (338 ridings - 120 Ontario ridings). Thus:
(338/(218))*.25 or ~38% of the vote, averaged across all non-Ontario ridings. I’d be interested in seeing where you got 2.7% from.
In an FTPT system, Ontario would have been swept entirely by Central, garnering 120 seats (35% of the seats for 25% of the vote). At 38% per riding, who could say how well Regional did? It could be an amazing showing if Orange and Forest Parties split the remaining votes very evenly, or it could be a horrid showing and they come up empty. So Regional could score from 0 to 218 ridings, a very random result for a very good showing.
Obvious issue: 36 Ontario ridings will have a non-Central MP (if any), despite Central having 64% of the vote.
Actually, it would retain its contitutionality. I’ve kept the same riding scheme. There are a few minor aberrations where some ridings only represent less than 20k voters, and there’s one riding with over 100k voters, but they average out to 75k each.
The MPs for a riding are selected locally as well. I’m well aware of the value of local politics and I relish having had the chance to speak with Ed Broadbent in person, even if we didn’t agree on all that much at the time.
The chief difficulty I have with this system I whipped up is that some MPs are chosen despite not having come close to winning that riding’s popular vote. This wasn’t too bad with most of the smaller parties (BQ and NDP had taken ridings with 25%-ish of the local vote). Except Green. The Greens took several ridings with less than 10% of that riding’s vote. Most of those ridings had also nominated a different MP as well, so were doubled up, and that was a more severe problem than I expected. Actual data was pretty illuminating.
The worst-case scenario would have been the Libertarian Party. In my 2015 simulation, they scored 0.7 seats but only ran 72 candidates. With another 30 candidates or so, they could have garnered a seat, despite only having a couple percent in any riding.
It’s not ideal, I know that. But I think it’d make a great pub chat over some pints.
I think most voters are bewildered by proportional voting schema, and making it as simple as possible was my goal. So here, a voter would simply show up and cast one vote for their MP and that was that, it was mechanically no different than our current FPTP system. Granted, that simplicity for the voters is revealing complexities in execution.
I don’t think you can get 25% of the votes across Canada by getting 2.7% of the votes in nine provinces and none in the tenth. Did you mean that the first party had a large majority in one province but no votes in any of the others, earning it 25% of the Canada-wide vote, while the second had about 25-30% in most provinces, also earning it 25% of the Canada-wide vote?
Most list-based proportional representation systems have a threshold to be granted list seats, which is usually around 5% of the vote or so. (Quebec’s proposed system has a threshold of 10% which is probably too high.) It seems unlikely that the population would accept an electoral system in which a party only has to earn 0.3% of the vote in each of Canada’s 338 ridings (actually less than that, since 0.7 seats as in this example should actually round up to 1) to elect an MP.
And as others have said, any system used to elect MPs to the Canadian House of Commons has to keep provincial representation in mind. So any list-based system would probably have to work with province-wide lists (which is not the case with your system). I’m not especially attached to local representation, though I understand that there’s a bit of an urban-rural divide there in the sense that especially for rural voters their local representative is often a useful resource. But provincial representation is definitely a constitutional requirement, not to mention a requirement for me.
Which it actually did by requiring parties to provide an MP for each riding, selected as they are now. The goal was to keep the process as identical as possible to our current FPTP system.
But that said, I can see how it would have been easy to miss that. I wrote my recent summary of conclusions in a disorganized flurry and the original explanation is several pages back in this thread.
Apparently, Hong Kong is indeed turning into an issue in the Canadian election. The conservatives are trying to paint Trudeau as too soft on China.
One person, one vote? In Canada it’s not even close
Edmonton-Wetaskiwin has 158,749 residents, while the electoral district of Labrador, has only 27,197 giving them almost six times the representation.
The numbers are a smidge different when you consider registered voters:
The riding of Labrador has 20,045 registered voters in 2015. Niagara Falls has 103,291, and Edmonton-Wetaskiwin is “second” at 100,871.
Source: Elections Canada, table E. I kind of agree with you but trying to get an MP to represent each district is kind of a pickle.
And in general, this election has felt like it’s been Scheer’s to lose, and if it weren’t for the NDP stealing the Liberal lunch, it likely would have been a Conservative loss. It may still turn out to be a win, eventually*. If memory serves, elections called on the failure of short-lived minority governments tend to favour the ruling party.**
- I’m feeling like it’s a Conservative minority, which the Liberals may dutifully prop up.
** Memory may not serve.