Proportional Representation for Canada:
How would small parties fare?
Reflections and two simulations
NOTE: The simulations in this post have been updated Sept 26 to reflect results as of Sept 24.
The results of this election once again demonstrate that Canada’s archaic voting system is fundamentally incapable of giving voters what they voted for. Results were wildly skewed and over 8 million voters (52%) elected no-one. Voter disengagement from the process is high: only 59% of eligible voters bothered to show up.
With these election results, many people are wondering how things would be different if Canada had proportional representation – and what would happen to small parties like the PPC and the Greens?
Frankly, many have expressed concerns about the PPC, a party that is widely perceived as far-right “extremist” and anti-vax winning a large bloc of seats in Canada’s Parliament with proportional representation.
When the media highlights simulations of purely proportional systems showing 20 seats going to the PPC, these fears increase. But those results are very unlikely in Canada. Our simulations (below) show why.
The role of politicians with “extreme” views in winner-take-all and proportional systems
All countries regardless of voting system have politicians with views that the vast majority find alarming or abhorrent, which still find support among some voters. What is considered “extreme” is quite subjective.
What is different between voting systems is how unpopular views are represented and how much power they have.
First-past-the-post hides the influence of those holding unpopular views, but doesn’t neutralize them. It’s crucially important to remember that Maxime Bernier lost by only 1.9% in the final count in the 2017 Conservative Leadership race.
An anti-vaxxer could easily be leading a party that under first-past-the-post needs only 30-something percent of the vote to form a “majority” government. That’s what happens when politicians with “extreme” views are in the big tent parties with first-past-the-post.
Proportional representation prevents minorities with extreme views from gaining control.
While there are always exceptions, in general, European countries with PR find increased voter choice helps isolate extremists. Governments must form coalitions, and they pay a heavy price for getting into bed with unpopular small parties. So those parties tend to remain on the outside, looking in.
For example, in Germany, the right-wing AfD party is in opposition in the federal parliament and all 16 state parliaments.
If a small party’s policies are too far out of line with mainstream opinion, the other parties will cooperate with each other and govern collaboratively, to ensure the smaller (extremist) party is excluded from government. This has happened with far right parties in Sweden, Germany, Spain and the Netherlands.
In short, proportional representation provides better representation for all voters. Often that means a couple of smaller parties win a few seats. But voters are better served because it’s very clear to all what those politicians stand for―extremists can no longer hide behind the scenes in one of the big tent parties.
But representation is not power. A party with 5% of the MPs cannot tell parties with 95% of the MPs what to do. As former Prime Minister of New Zealand Helen Clark pointed out, the tail can’t wag the dog. Rather, the dog has a choice of which tail to choose.
Also, as our simulations show, the models of PR recommended for Canada make it easier for smaller parties to win seats, but not the big blocks of seats the media is showing right now. To get into the nitty gritty of what the results of the 2021 election might look like if votes had been counted using commonly recommended PR systems, read on….
Two simulations using made-for-Canada of models of PR
To answer the question of how the PPC and the Green Party might have fared with PR for Canada, our experts have prepared two simulations based on how people voted in this 2021 first-past-the-post election.
Two of the systems most commonly recommended for Canada have been used: Mixed Member Proportional (MMP) and Single Transferable Vote (PR-STV).
It’s important to note that no model of proportional representation suggested for Canada is strictly proportional. To accommodate our geography and maintain strong local representation, they tend to be designed to deliver moderately proportional results.
At the end, we’ve also provided real-life examples for both of these systems of a small party, from countries with systems similar to those used in our simulations.
Caveat: When you change the voting system, voter behaviour changes, too, so real-world results would no doubt differ from our simulations. With proportional representation, people feel less pressure to vote “strategically” for the lesser evil. Research also shows that voters are less likely to vote for far-right parties in proportional systems compared to those with winner-take-all systems like ours.
Single Transferable Vote (PR-STV)
(Full details see: http://election-modelling.ca/2021_static/2021/stv_med/index.html)
With Single Transferable Vote, several local ridings are merged and voters collectively elect several MPs using a ranked ballot (allowing them to mark candidates 1, 2, 3 etc. in order of preference). STV is a great example of how a ranked ballot and proportional representation can work together.
If moderately proportional STV had been used in the 2021 election, the results would have been more proportional:
Note that the seat results for most parties are much closer to proportional than Canada’s current system, but not a perfect match.
For those noticing only one seat for the Green Party (compared to 2 they just won with first-past-the-post), it’s important to note that the Greens just had their worst election in decades.
There were no Green candidates on the ballot in one quarter of the ridings in Canada, so those votes weren’t there to factor into a simulation. Greens received only 2.3% of the national vote―about one third of their share in 2019. A simulation cannot capture an outlier like Mike Morrice’s win in Kitchener Centre.
When their support bounces back to more traditional levels, we are confident that Greens would win substantially more seats under STV than first-past-the-post, as they do with STV in Ireland, Tasmania and Australian Capital Territory.
Mixed Member Proportional (MMP)
(Full details see: http://election-modelling.ca/2021_static/2021/mmp_small-MMP_FPTP/index.html)
With Mixed Member Proportional, voters elect both a local MP in a larger riding using first-past-the-post and one or more MPs that serve the entire region. The regional MPs, selected by voters from an open list, compensate for the distortions of the first-past-the post results and ensure that the overall seat totals for each party more closely reflect their share of the popular vote.
If a moderately proportional MMP system had been used in the 2021 election, the results would have been:
Two real-world examples showing how small parties fare under PR-STV and MMP with designs similar to our simulations
Both of the examples below demonstrate how small parties fare in moderately proportional systems where the barrier to gain seats is lower than first-past-the-post, but higher than in perfectly proportional systems. Similar to the models for Canada described above.
Single Transferable Vote (PR-STV): The Green Party in Ireland
With PR-STV, Ireland elects small teams of local MPs in multi-member ridings using a ranked ballot. The ridings elect 3-5 MPs each.
Having several local MPs means that the different points of view in each local community are represented in the legislature. In other words, proportionality is built in from the ground up.
In a 5-member riding, a candidate needs about 16.7% support from local voters to get a seat, although the last candidate can be elected with substantially less.
How does this work out for voters of small parties?
Ireland regularly elects representatives from smaller parties in areas where they have popular local candidates.
In 2016, the Green Party got 2.5% of the popular vote in Ireland, but only 1.2% of the seats.
They were represented by two of their strongest local candidates.
A fully proportional share of seats for their party would have been about four representatives.
In 2020, the popular vote for the Green Party rose to 7.5%. In that election, they secured 7.7% of the seats (12 representatives). Fully proportional.
The barrier to election is much lower for small parties than under first-past-the-post, giving them some representation. However when a party is very small, the results are not usually fully proportional.
Mixed Member Proportional (MMP): The Green Party in Wales
In Wales, voters elect local representatives (riding MPs) by first-past-the-post and regional MPs from party lists. Each region contains 12 seats: 8 first-past-the-post seats and 4 proportional members for regional top-up seats.
To win a proportional list seat in one of the regions in Wales, the party needs to get about 8% of the vote in that region.
In 2021, the Green Party won 4.39% of the popular vote, but failed to win any seats. They didn’t win any ridings with first-past-the-post and their level of support was too low to win any proportional list top-up seats.
In that same election, another small party, the Liberal Democrats got 4.3% of the vote but they elected one MP in the 60 seat chamber. This MP was elected in a riding using first-past-the-post, just like our Green Party wins seats in Canada. The party failed to win any proportional list seats.
In other words, with a moderately proportional MMP system, smaller parties can break through much more easily than with first-past-the-post, but they must attract enough of the popular vote in a region to qualify for proportional list seats.