Some lessons from Canada's most recent election (and the previous one).
With apologies to my non-Canadian readers this time I am going to do a deep dive into Canada's electoral system. Just a few days ago we had an election where it turned out we dodged a bullet by not allowing the Conservative Party of Canada, Canada's answer to Donald Trump and the Republicans, to form the government. This came at the expense of losing representation in Parliament of the left, which is a serious loss for Canadian society in my opinion, and to make matters worse it is, as I show below, an artifact of the electoral system.
What if the different political parties were represented in Parliament in proportion to the number of votes they got? This is called proportional representation. It implies that everyone's vote has exactly the same weight. This is only fair, which is why it is also called fair voting. Fair Vote Canada is a national organization devoted to promoting the idea of fair voting. They have a lot of information which you may find interesting.
More specifically we can ask how many seats each party would have ended up having if the number of seats was proportional to the number of votes. This is not a difficult calculation, and the result, which I will show you in a moment, is very revealing. There is a subtle little difficulty, which is how do you assign a small integer number of seats in proportion to very large numbers of votes. It can't be done exactly, but there is a way to come as close as possible, called the D'Hondt method. I wrote a little Python notebook to calculate the seat distribution using this method. (Let me know if you would like a copy; my email is in the Contact Me page. It's very easy to adapt to any parliamentary election, just plugging in the relevant numbers). Elections Canada hasn't yet published the numbers, but I found the number of votes each party had on the Global News website, except for the People's Party of Canada which I found on Wikipedia. I haven't included other parties or independents, assuming that their numbers would be too small to alter the results significantly.
So <play drum roll here> here are the results:
| Party | Votes | Seats obtained | Seats with PR | Difference |
|---|---|---|---|---|
| LPC | 8,565,147 | 169 | 152 | 17 |
| CPC | 8,089,420 | 144 | 143 | 1 |
| BQ | 1,233,004 | 22 | 21 | 1 |
| NDP | 1,237,446 | 7 | 21 | -14 |
| GRN | 244,963 | 1 | 4 | -3 |
| PPC | 141,210 | 0 | 2 | -2 |
A couple of things jump out immediately: The Liberal Party got 17 seats more than they should have, and the NDP got 14 seats fewer than they should have; and the Bloc and the NDP should have got the same number of seats, having got practically the same number of votes, and yet the Bloc got roughly $\pi$ times as many seats as the NDP. Had the NDP had its proportional share of the seats, there would be no question about losing party status.
The comparison between the Bloc and the NDP illustrates perfectly a major flaw with First Past The Post: it is heavily biased towards geographical concentration of votes. The Bloc only runs candidates in Quebec, with about 6.5 million electors, while the NDP competes nationally, with about 37 million electors. Another extreme example is the Green Party. They only won a seat because Elizabeth May is very popular in her own riding, but the rest of their supporters are spread too thinly throughout the country to be able to win a single seat.
OK, you say, what about the previous election in 2021? It's very easy to plug in the numbers into my little program, and the results are even more shocking.
| Party | Votes | Seats obtained | Seats with PR | Difference |
|---|---|---|---|---|
| LPC | 5.556,629 | 160 | 112 | 48 |
| CPC | 5,747,410 | 119 | 115 | 4 |
| BQ | 1,301,615 | 32 | 26 | 6 |
| NDP | 3,036,348 | 25 | 61 | -36 |
| GRN | 396,988 | 2 | 8 | -6 |
| PPC | 840,993 | 0 | 16 | -16 |
OK, first of all: The LPC got 48 more seats than they should have? I am no fan of the Conservatives, but can you imagine getting more votes than the Liberals and 41 fewer seats?
And then, the NDP. They got nearly two and a half times the number of votes of the Bloc and three quarters the number of seats! At 61 seats for the NDP and 112 for the LPC, the NDP could certainly have demanded a coalition with seats in cabinet, and not a very grudgingly conceded small confidence and supply agreement.
As for the Liberal Party, it is very clear from both elections why Trudeau was willing to pay the political price of promising to end First Past The Post, and then breaking his promise as the parliamentary committee on electoral reform held public hearings and received submissions from all over the country, overwhelmingly in favour of proportional representation. The Liberal Party gains the most from FPTP, and the NDP, their rivals on the progressive end of the political spectrum, loses the most from FPTP. Trudeau's preferred method of instant runoff is not proportional, and in fact produces very similar outcomes to FPTP.
This calculation shows the failings of FPTP from the point of view of the parties. There's another simple calculation we can do that shifts the spotlight onto the voters themselves: let's calculate how many votes it takes to elect an MP for the supporters of the different parties. This simply requires dividing the second column by the third in the tables above. Here's the result (not including the PPC because they didn't win any seats):
| Party | Votes per MP in 2025 | Votes per MP in 2021 |
|---|---|---|
| LPC | 50,681 | 32,879 |
| CPC | 56,177 | 39,913 |
| BQ | 56,046 | 59,164 |
| NDP | 176,778 | 433,764 |
| GRN | 244,963 | 396,988 |
The violation of the principle that everyone's vote should be equal is painfully evident. It takes over three times as many voters to elect an NDP MP compared to a Liberal or Conservaitve MP; and it takes nearlly five times as many Green voters to do the same. It's also interesting to note that from the point of view of the voters we don't have a two-party system as we often hear, but a three-party system. The Bloc has been able to buck the natural trend of FPTP to produce just two big parties by only running in their home province of Quebec.
Canadians have been discussing electoral reform and fair voting for a long time. Notably the Law Commission of Canada, an arm's-length commission designed to advise Parliament on law reform, released a report on electoral reform titled “Voting Counts: Electoral Reform for Canada” in 2004, after two years of research. In it they proposed mixed member proportional representation, a system used by, among others, Germany whereby voters get two votes, one for their local representative and one drawing from a party list to top up the number of seats for each party so that they are proportional to the number of votes. This system satisfies both the requirement of proportionality and that of local representation. Stephen Harper shut down the Commission two years later.
It is high time that we abandon the blatantly unfair First Past The Post. We cannot call Canada a democracy until we do.
#Canada #ProportionalRepresentation #FPTP #Voting #Democracy