It can make you ask, “What’s the point?”

by Maurice Y. Michaud (he/him)

My niece Julie was living in a very politically conservative part of British Columbia in 2011, when Stephen Harper finally formed his majority government. I think she found me a little bit obsessed in the weeks leading up to that election, but to me the fact that the Harper government had been found in con­tempt of Parliament was reprehensible and meritorious of punishment, not reward. Even though she lived in a riding where her one vote would get lost in a sea of votes opposing hers, I still urged her to go vote. So the day after, with the Conservatives having indeed been rewarded with a majority, she commented on my Facebook page, “Well I didn’t vote for them but they got in anyway.”

Implying... what’s the point?

A majoritarian electoral system like FPTP can make anyone feel that way. A riding where three candidates are competitive can be won with a third of the votes, with Nobody being the real winner. Conversely, in a riding where the support for a given party is consistently overwhelming, like this one, one can almost be forgiven for not bothering to voice their dissent. Another majoritarian system, Instant runoff or alternative voting, gives the impression of a “real” winner within a riding but leads to more distortion than FPTP in the final seat counts. In short, one can feel that not only their vote but most people’s votes get lost and don’t matter.

By way of explanation (if not outright excuse), political scientists and analysts have concocted the notion of “vote efficiency” to explain how majoritarian systems work. To achieve vote efficiency, parties must examine closely the demographic profile of each riding to identify those where their platform is more likely to appeal to its electorate, who can help them get first past the post. Therefore, given that to win, candidates need only one vote more than their closest rival, parties need to target as many ridings as possible where they have a reasonable chance of reaching that baseline. That’s because what matters in a majoritarian system is not winning votes, but winning seats.

As a result of compiling the results of all the general elections held since Confederation, I have come to make a distinction between lost votes and unperforming votes. That distinction is stated on the summarized statistics of each election’s detailed results.

Lost votes are defined as being “Votes that did not contribute to a party winning at least one seat, plus the votes for a ban­ner (or non-party) like Inde­pend­ent that did not yield to winning a seat.” Un­per­form­ing votes are those that did not contribue to winning a seat, while those that did are deemed performing. Because the vote efficiency of each party can vary by region, the sum of the lost votes from each region only adds up if the party has won no seat (in which case “lost” and “unperforming” are equal), while the sum of the performing or unperforming votes from each region always adds up.

Julie didn’t mention to me who she ended up voting for, but given the choices she had, her vote wasn’t lost; it was unperforming.
 

Distinguishing between the two types

To understand the difference between lost and unperforming votes, let’s begin by setting aside marginal parties and independents. That’s because in a well-designed proportional system, marginal parties wouldn’t stand a chance of reaching a critical mass of support, while independents in partisan assemblies are lone wolves who generally eschew party affiliations.

In a general election, the number of unperforming votes is always much greater than the number of lost votes, so does that make the former worse than the latter? Arguably, no. Lost votes are much worse for political parties because they are as if no one had voted for them, whereas un­perf­orm­ing votes are as if not enough people had voted for them. That being said, it doesn’t take much shifting in the final results to turn votes from lost to unperforming — sometimes with fewer votes from one election to the next. That’s where the notion of vote efficiency comes in.

Green Party of Canada

The Green Party of Canada in the 2008 and 2011 general elections provides a striking illustration of this difference.

Winning one seat nationwide in 2011 with fewer votes overall made a world of difference for the Greens, as it eliminated their lost votes but far from solved their problem with representation. With 308 seats in the House of Commons at the time, one seat as a percentage equalled 0.32%. In a purely proportional system (for which PoliCan does not advocate), one might have expected them to win 11 more seats. However, in a mixed pro­por­tion­al system, by having obtained only 3.89% of the votes across the country — far below their highest watermark of 6.78% reached in 2008 — they unlikely would have met the minimal threshold to obtain regional (or compensatory) seats.

Now, let’s read again slowly the last sentence of the statement on the summarized statistics of each election. “Because the vote efficiency of each party can vary by region, the sum of the lost votes from each region only adds up if the party has won no seat (in which case “lost” and “unperforming” are equal), while the sum of the performing or unperforming votes from each region always adds up.”

To help you wrap your head around this statement, let’s take a look at the same results plus those from 2019 and 2021, but let’s break them down by regions: West (& North), Ontario, Québec, and Atlantic. (I’m skipping 2015 because the Greens again won only one seat in B.C. that year.)

A simple way of understanding the distinction (and these numbers) would be to say:

• Lost votes: What was the point of voting for them?!
• Unperforming votes: Better luck next election!

Progressive Conservative Party of Canada

The spectacular collapse of the Progressive Conservative Party of Canada provides another good illustration. Given how in 1993 that party went from a governing majority to only two seats in the Commons, it would be tempting to say that almost everyone who voted for that party that year lost their votes. But it would be more accurate to say that these people’s votes had become terribly unperfoming to the point of being lost in some regions. To fully grasp this observation, let’s begin by looking at the results of the PCs in the general election preceding their ’93 debacle up to the last general election in which they took part.

Despite a modest bounceback in 1997, the federal Progressive Conservatives were simply never able to get their votes to perform again. So, after 2000, they — and especially their conservative rivals — could see the proverbial writing on the wall, and a decade after their brutal spanking of 1993, the PCs — the party of John A. — ceased to exist.

The 2024 Nova Scotia general election

The divide between urban and rural ridings has always existed in Canada. However, it has become more stark in the 21st century, especially at the provincial level. This has been the case for many years from British Columbia to Ontario, but as of 2024, it has also become the case in Nova Scotia. That comes to the chagrin of the NDP, which formed government only once in its history, but particularly for the Liberals, who have governed the province more than any other party. But of course, the Progressive Con­ser­va­tives couldn’t be more pleased!

This tabulation looks at the results of the 2024 general election for all the parties as well as the independents, by region, through the lens of total, performing, un­per­form­ing, and lost votes. The regions are defined as the Halifax Regional Municipality (Hfx), Cape Breton Island (CB), and the rest of Nova Scotia (Rst).

Observations (including something that’s not shown on this table):

1. Not shown: This election had the lowest voter turnout in the province’s history, including the war years.
    
2. Looking at the province as a whole, it would seem that only 3,593 votes were lost. But after breaking down each party by region (with special treatment for “Independent,” which isn’t a party), it can be argued that the number of lost votes was in fact 55,963 (2,055 for Halifax, + 302 for Cape Breton, + 53,606 for the rest of the province). Looking at these numbers in this manner is more realistic, as it better takes into account how some rural areas have only alternated, to a greater or lesser extent, between the Con­ser­va­tives and the Liberals, and that the votes for any other party have always been so unperforming that they ended up being lost votes.
    
3. The NDP clearly had a better understanding of vote efficiency than the Liberals in this election, for they won 7 more seats than the Liberals with 1,844 fewer votes than them.
    
4. If, in addition to the other two seats that were won, the Liberal leader had won his seat, which happens to be classified here under “Rst,” or had the incumbent in a nearby riding also classified under “Rst” not lost his seat by 8 votes, the Liberals’ number of lost votes would have been 0. However, their unperforming number would still have been through the roof.

In short, the impact of the more accentuated urban/rural cleavage in the provinces is that the votes for some parties have become much less per­form­ing — in some cases to the point of getting lost, as was the case for the NDP and the Liberals in the rest of Nova Scotia in 2024.

The 2022 Québec general election

This last tabulation helps us understand what happens when one party is only relatively dominant but benefits from having a fragmented opposition. It also lends credence to the notion that breaking down results by party and regions reveals a more accurate number of lost votes.

Particularly remarkable in the 2022 Québec general election was how the Coalition avenir Québec was able to form an overwhelming majority of 72% of the seats with 41% of the votes provincewide. The Liberal Party, thanks to its vote efficiency in Montréal, finished second in the seat count but fourth in the popular vote. Québec solidaire was the party coming second in the vote count (thanks to its strong support on the populous Island of Montréal), but third in the seat count. The Parti québécois was the party coming third in the vote count, but a distant fourth in the seat count. And the reinvigorated Conservative Party, despite garnering just shy of 13% of the votes, was shut out of the National Assembly. The remaining 1.7% of the votes went to 256 candidates representing 16 other parties or as independents.

You’re forgiven if reading that paragraph made you dizzy! But imagine how everyone in Québec felt the day after this election, as they tried to understand the following numbers.

Observations:

1. Again, if we look at the province as a whole, it would seem that “only” 600,928 votes were lost. But this analysis finds 844,771 more lost votes, or 1,445,699.
    
2. The CAQ shut out all opposition in two regions: Laurentides & Lanaudière, as well as Saguenay, Lac-Saint-Jean, North Shore & Northern Québec, respectively with 49.9% and 54.9% of the popular vote. That means that half or nearly half the votes in those regions were not just unperforming but outright lost.
    
3. LaFontaine was the Liberals’ easternmost seat, but that riding is on the Island of Montréal, which is in the southwest corner of Québec. Yet due to an abundance of seats in southwestern Québec, combined with their vote efficiency in the Montréal area, they came in second in the overall seat count while only being fourth in the vote count.
    
4. Despite getting almost twice as many votes than Québec solidaire in the Capital Region, Charlevoix & Chaudière-Appalaches, the Conservatives won no seat while QS won two.
    
5. Also, with a provincewide vote share less than 1% greater than what the Parti québécois obtained, QS won 8 more seats than the PQ.

Quite frankly, aside from referring to “vote efficiency,” none of the results made much sense, but we had FPTP to thank for that!
 

So “what’s the point” of this distinction?

Eighty-nine percent of the 1,054,395 votes deemed lost in the jurisdiction as a whole in the 2008 federal election, and roughly that same percentage of 600,928 votes in the 2022 Québec general election, had been cast for a single party (the Greens in 2008 and the Conservatives in 2022). Such numbers should, by themselves, be raising a red flag about our majoritarian electoral system rather than be dismissed as mathematical flukes. Moreover, the fact that parties must pay so much attention to vote efficiency is also a problem, as it is a notion that is disconnected from the interests of the voters and distorts what they are trying to communicate through their ballot.

Therefore, the point of this distinction is that it helps to pinpoint how a proportional electoral system would go a long way towards rectifying problems that are inherent to a majoritarian electoral system.

• In the case of the Green Party in 2008: If the threshold had been set to 5 percent of the votes nationally, the votes it received would not have been lost. They would have been unperforming and some of them could have been compensated.
   
• In the case of the federal Progressive Conservatives in 1993: If their unperforming votes had been given a value instead of being ignored, that party’s demise may have at least been postponed, if not prevented. Consequently, the impact of unperforming votes would have been mitigated, as the votes given to parties reaching the minimal threshold would be considered in the allotment of the compensatory seats.
   
• In the case of the growing urban/rural split within the provinces: It would prevent forcing NDP supporters living in Annapolis from having to move to Halifax in order to have their voice heard. No but, joking aside, if lost votes in rural or urban regions could be converted to un­per­form­ing votes, the impact of those localized unperforming votes could be mitigated.
   
• Finally, in the case of Québec in 2022: Dismissing it as an anomaly would be disingenuous and indicative of electoral amnesia. We only need to go back to 1973 to find another set of crazy results. That is when the Liberals won 93% of the seats (102 of the 110 at the time) with 54.7% of the votes, leading to a phenomenal number of lost votes.

In a proportional electoral system, the most egregious type — lost votes — would be eliminated for parties obtaining a reasonable amount of support, while a good number of unperforming votes would be given a commensurate value rather than all being thrown in the trash. The notion of vote efficiency would become passé, and the notion of winning votes rather than seats would be restored. Plus, an added benefit would be that strategic voting, which unfortunately few voters understand where, when and how to apply it, would no longer be necessary.

So that’s the point of me making this distinction, and that’s why it’s important for us to understand it.

---