-
There is a system that is getting a fair bit of buzz up here in Canada. It is an older system but people seem to think it is viable. It is called Single Member District Proportional Representation. Watch the video for details.
It is basically Party List but you vote like Single Member Plurality so that the list can be determined by the fraction of vote they get in their district. I think this makes it better than Party List in a clear way. However, Party List is the worst system (IMO) so thats not much. It is also similar to the second part of Dual Member Proportional which I think is better so I do not see the point. The only thing one could claim DMP lacks compared to SMDPR is a one to one relationship to the voters but I am not sure SMDPR can really claim that given it is not a proper single member system.
I would think that somebody could vastly improve this system with a score ballot and get rid of all the vote splitting and clone issues. The list would be derived via the highest average score and the seat allotment could run any Cardinal PR method. Sorting out conflicting seats might need some work but I am sure it is doable. That would likely be a system better than MMP. It still hurts independents and is pretty complex but might be viable.
Although.... at that point we are basically at running a proper multi-member cardinal PR system but with Local district clusters. So why not just do that. What you lose there relative to SMDPR is the level of PR but with 5 member districts people tend to be happy with the level of PR so I am not sure there is an issue there.
Am I missing something?
-
@keith-edmonds I think it will be difficult to sell having a candidate represent a district where they didn't win. There could also be extreme scenarios that make this more unpalatable. Suppose the Green Party barely clears a 5% threshold in Ontario (for example, in the 2019 federal election, they got 6.2%, which would entitle them to about 5 seats.) In 2019, the Greens did not do particularly well in any individual Ontario riding. To get them to 5 seats some of them would have to be put in ridings where they got something like 15% (possibly lower if some of those get assigned to a different party.)
-
@marylander That is not something I am super concerned about because this happens in all PR systems. This system just makes it more clear how PR is traded for local representation. Although, it could be that this system does not balance this trade-off well. I am not sure how that is measured but in a system like Party List where nobody has a local representative the harm is uniform. In this system that does not seem to be the case.
@Marylander What are your thoughts on this vs a Cardinal PR system? Are there any advantages other than simple ballots? The negatives as I count them are
- Low ballot Expression
- Difficult to explain
- Hurts/eliminates independents
- Imbalance of lowered local representation
- Province wide tabulation needed
- I also suspect Duverger's law would still basically apply making this a polarizing system
-
@keith-edmonds In general, I think it can actually be a disadvantage to have a new system look too similar to the old one because some people might not notice the difference, although good communication by the election authorities should prevent this. This system seems intended to minimize changes in the presentation of the vote to the voter (5 minute mark of the video).
I don't think Duverger's law will apply since I think voters will essentially vote as if it was party list, since winning pluralities in districts doesn't matter.
@keith-edmonds said in SMDPR:
That is not something I am super concerned about because this happens in all PR systems. This system just makes it more clear how PR is traded for local representation.
I would still expect this point to come up if there was ever a referendum on this system.
-
This post is deleted! -
There is no need to increase the number of MP's as you would in DMP? SMDPR is a proper single member system as each riding gets one representative who ran there. SMDPR as I have it includes independents. In 2019, 2 independents were elected according to my limit of 5% of the vote in each riding. If less than 5%, the candidate with the highest vote count wins. As I run it, there is no party limit as to the percentage of the vote - just in the ridings.
In Ontario in 2019, SMDPR gave the Greens 8 seats all of which had at least 5% of the vote in their ridings: 25.99, 25.46, 10.86, 7.50, 7.06, 6.52, 6.29, & 6.26%.
-
Also, electoral districts or divisions can be less than 15 ridings - the bigger provinces do not need to be tabulated all at once.
-
There is no need to increase the number of MP's as you would in DMP?
This is not a requirement in DMP.
SMDPR as I have it includes independents.
How could an independent win? They would have to have nearly all votes in the riding. They are a best facing a huge disadvantage.
In Ontario in 2019, SMDPR gave the Greens 8 seats all of which had at least 5% of the vote in their ridings: 25.99, 25.46, 10.86, 7.50, 7.06, 6.52, 6.29, & 6.26%.
What was the % of support for the FPTP winner who was displaced by a green with less than 10%?
What do you think about doing a more sophisticated version of this with score ballots or something?
-
I never heard of DMP until I read your comment on it. I looked into it for a minute and it seemed as if it required 2 people per riding.
An independent won in BC because the PPC candidate that was picked only had 0.81% of the vote so the independent had the highest vote count of 32.65%. The winner in Ontario was picked as the final pick in the province. She happened to be the last one left and had 20.72% of the vote.
If an independent gets a lot of votes along with a lot of other independents running in the region or division they stand a good chance of winning.
The % support of the displaced candidates was 39.53 for 7.50, 39.25 for 7.06, 40.55 for 6.52, 45.89 for 6.29, & 45.62 for 6.26%. Those displaced were all Liberals whose SMDPR ranked list was quite short compared to their FPTP winners' list.
Both Dr. Threlfall and Dr. Falvey have suggested changing the ballot that would exclude independents. I don't know much about score ballots but am afraid they might exclude independents as well. If not, I'm open for change.
The other reason I don't like changing the ballot is that a lot of voters might have difficulty understanding how to vote.
Changing the limit on vote percentage might be alright too. But if you have divisions with less than 15 ridings the percent required for a party to have a candidate picked approaches 5%; and if 10 or less the percent is at least 5 to have a winner. For instance, in PEI with 4 seats, the percentage for a party to get a winner is 12.5
Dave.
-
I looked into it for a minute and it seemed as if it required 2 people per riding.
You can also just double the size of the district.
The % support of the displaced candidates was 39.53 for 7.50, 39.25 for 7.06, 40.55 for 6.52, 45.89 for 6.29, & 45.62 for 6.26%.
All PR systems do this in one way or another but this is much more in the open. I wonder how public opinion on that will be.
I don't know much about score ballots but am afraid they might exclude independents as well. If not, I'm open for change.
The information you put on the ballot is independent of the candidates. Score would actually reduce vote splitting so help independents in the end.
The other reason I don't like changing the ballot is that a lot of voters might have difficulty understanding how to vote.
Score is just giving a rating. People do this all the time. There are studies that show it is faster for people to rate a bunch of things that choose their favourite. Ranking is the slowest.
-
Sounds good. I have thought over the problem of cutting out too many candidates with a large percentage of the vote in their ridings. Maybe re-programming my computer again to limit the party percentage to 5 and limiting the riding percentage to 10 will help people be more acceptable to the system although it would restrict smaller parties and independents.
