Proportional Representation: Decisions, Decisions
Last week, I wrote about Prime Minister Trudeau’s preferred method of electing MPs to the House of Commons, which is the Instant Runoff Vote. However, there is a growing call for Proportional Representation (PR), where a party would get a share of MPs that roughly mirrors their share of the vote.
But it’s not that simple.
The European Parliament serves as the legislative body for the European Union. Each of the 28 member states gets a certain number of MEPs to send to Brussels and Strasbourg, and each country has to use PR to elect them. However, each country has slightly different rules for how MEPs are allocated to their respective parties, and these subtle differences can have a considerable impact to the overall outcome.
Germany elects its 96 MEPs using the purest form of PR, with the national vote being used and no minimum vote requirement. With a 338-seat parliament, a party would need about 0.3% of the vote to enter Parliament. While only the five main parties reached this last year, one other (the Libertarians) came close, and others may have as well had it not been for the tactical voting that comes with First Past the Post. A bar as low as 0.3% could mean more parties in the House, and thus make passing legislation more difficult.
Next door in Poland, their 51 MEPs are allocted amongst the parties that get at least 5% of the vote nationwide. This excluded three parties which otherwise would have won seats, and if applied in Canada, it would lead to the exclusion of the Bloc Québécois and the Green Party. Politicians will argue that dealing with fewer party leaders would lead to a more effective government, but with these two parties having quite concentrated support (in Québec and BC, respectively), voters in those provinces are likely to feel disenfranchised.
The United Kingdom’s 73 MEPs are divided between the country’s 12 regions, and then are allocated proportionally within each region. This has an impact of having a higher effective threshold to get anyone into Parliament. If applied to Canada, assuming no change to how MPs are allocated to provinces, this will lead to effective thresholds ranging from 0.8% in Ontario to 20% in PEI. The biggest impact will be experienced by the Greens, who are only able to get seats in the three largest provinces (and those with the lowest effective thresholds), and they’d lose about a third of their seats as a result. The Liberals make big gains from this approach, though, largely due to the fact that the Atlantic provinces, where they did extremely well last year, have more seats than their populations imply they should.
Across the channel from the UK, France’s 74 MEPs are both divided by region and subject to a 5% threshold within each region. This kind of threshold would impact four provinces which don’t have effective thresholds already above this. This would exclude the Greens from all provinces bar one, surpassing the 5% threshold only in British Columbia. The Conservatives, NDP and Bloc are all within a seat of what they’d get under pure PR, while the Greens would lose 2/3 of those seats and the Liberals gaining a significant share.
Finally, we go to Ireland, where we see the Single Transferable Vote being used to elect its 11 MEPs. Here we divide the country into ridings with 3–5 seats each, with the aim to go approximately-proportional within each riding. To do well in this kind of system, you would still need to have concentrated pockets of support, like the Liberals have in Toronto and the Tories in Alberta. Parties whose support is spread out, like the Greens and the NDP, will do almost as poorly under this system as under First Past the Post.
Proportional representation does have its pros and cons, as do each of the subtle differences in approach that I’ve been through. The committee looking into changes to our electoral system will need to consider the impact that each of these decisions will have, not just on the outcomes but on how the public will feel about them.
Note: With the exception of STV, all of these estimates use d’Hont’s Method to allocate the seats between the parties. Other methods are available, but most of these have fundamental flaws.