Tuesday, June 26, 2018

Election Reform - Ontario under different systems

There are a number of systems we can look at, lets start with some of the simpler ones

Alternative Vote (Instant Runoff Voting)

The Liberals would lose their seat in Thunder Bay, but gain Eglinton-Lawrence.

The Tories would not gain any seats. The Tory loss in Eglinton-Lawrence to the Liberals is offset by the NDP gain from the Liberals in Thunder Bay, meaning we begin the counter with the NDP up by 1.

The NDP further takes the following ridings:

The soo
Ottawa West
Rouge Park
Brampton West

and quite possibly also

Scarborough Centre
Etobicoke Lakeshore
and Brantford Brant

For a total of 12 NDP gains from the Tories, pushing our numbers to a final result of:

64 PC
52 NDP

A narrow PC Majority. Note as well that the last three ridings I listed were quite iffy on the math, and its quite likely that in the event the real election had used IRV, these three would have remained PC, pushing our split from 64-52 to 67-49, a much more comfortable majority.

Pure PR (Proportional Representation)

A pure PR system would see a single province-wide "riding" that elects 124 candidates. According to the popular D'Hondt method of counting (and the method whose online calculator is most user friendly in my experience) the results are as follows:

51 - PC
43 - NDP
25 - LIB
5 - GRN
0 - LBT

The Libertarians fail to win a seat even without a threshold.


This post originally was going to examine STV and other systems; and maps have been prepared for that purpose. However, other bloggers are dealing with STV, and as such, I've decided to truncate this post and end with an examination of a Parallel PR system.

Parallel PR

A Parallel PR system mixes the benefits and drawbacks of both PR and FPTP. Unlike a supplementary or "fill up" system, a parallel system is additive.

Longtime readers will know this is my reform of choice for Canada and its provinces. Two arguments seem to carry the most weight in Canadian debates about electoral reform. First, people want to be able to keep majority governments, even if they are, as some in the NDP have tried to classify them "false majorities". And second, Canadians are concerned about representation of the opposition; first that the main opposition party can, often, find itself without representation from certain entire regions, and secondly, that the third party and lower, can find itself with very few, if any, elected representatives. Parallel deals with both of these in ways Canadians want to see.

I also want to note that a true Parallel PR system would not see the total number of seats change much, but that I am using a larger number of seats as this saves about 6-8 hours worth of calculation of manually re-drawing every riding, and makes the calculation time for the Parallel system total around 2 minutes for calculating how popular vote turns into additional seats won. I use this d'hondt calculator for that purpose.

Additionally, most of the systems I propose for Canada have only a small number of PR MPs, as, polls do show Canadians are iffy on the concept of a mixed-member system, and having a smaller number is more likely to succeed. As such, to the 124 ridings, I am adding an additional 24 proportional seats.

Of these seats, the PC party wins 10, the NDP wins 8, the Liberals win 5, and the Greens take 1.

As such the final end results are as follows:

86 - PC
48 - NDP
12 - LIB
2 - GRN

As an example of how this may play out in reality, we could expect to see the 10 new PC MPPs from all over Ontario, including London, Hamilton, Toronto, Ottawa, Northern Ontario, and various more rural areas. The NDP is more likely to see members elected from Mississauga, the York Region, and various rural areas, while the Liberals could easily end up with representation not just from Toronto and Ottawa, but in places like Windsor, Kitchener, Mississauga, and both rural eastern and southwestern Ontario. This would mean that all 3 parties have members from Mississauga, for example; and given how Parallel systems are usually implemented, this is extremely likely.

No comments:

Post a Comment