The Pendulum

A government fell further on Monday than any provincial government in Canadian history has fallen. That sentence is true, and it is the wrong place to start. The interesting number is not ninety to zero. It is this: the Coalition Avenir Québec collected 571,520 votes and won no seats at all, while Québec solidaire collected 532,176 votes and won nine. Thirty-nine thousand more votes, nine fewer seats, in the same election, under the same rules, counted honestly. Nothing went wrong. A machine sat between the ballots and the seats, and it did what it is built to do: it converted. This Edition reads the converter. The vote is real and the count is real. What is laundered is alternation, the swing itself, re-presented as the public being heard, when the size of the swing is set by the gearing and not by the public.

§01 · What the machine returned

Start with the record, bare. These are the preliminary results published by Élections Québec for the general election of 5 October 2026, with all 20,704 polls reported in all 127 ridings.verified The National Assembly grew from 125 seats to 127 for this election, under the final electoral map published in the Gazette officielle du Québec on 14 January 2026.verified

4,282,551 valid votes · 63,306 rejected ballots · turnout 66.89% · 6,495,984 registered electors

Four of those columns are facts about the public. The fifth is a fact about the machine. A party that ran 127 candidates and finished second in nineteen ridings holds no seat in the legislature; the closest it came was Jonquière, where its candidate took 33.34% and lost to 36.34%.verified There is no rule anywhere in Quebec law that says a party with 13% of the vote gets nothing. The rule says only this: in each riding, the largest pile wins and every other pile is discarded. Apply that rule 127 times and the discarding adds up into a shape nobody chose and everybody then reads as a decision.

§02 · The price of a seat

The clean way to see a conversion is to put the input beside the output and ask what the exchange rate was. Here it is, for one election, on one chart.

The exchange rate is not a policy anyone defended. It is a residue.

The single comparison worth holding is the fourth row against the fifth. The Coalition Avenir Québec finished ahead of Québec solidaire by 39,344 votes across the province and finished nine seats behind it.verified The difference between them is not popularity. It is geography: Québec solidaire's support is stacked in a small number of urban ridings where stacks win, and the Coalition's was spread evenly across ridings where it came second. A system that counts piles rather than people will always pay the concentrated more than the numerous, and it will never say so on the ballot.

§03 · The same gear, four years apart

If the 2026 conversion were an accident, the previous one would look different. It does not. It looks like the same machine running the other way.

In 2022 the Coalition Avenir Québec took 40.98% of the vote and 90 of 125 seats: 72% of the legislature, a crushing majority built from two votes in five.verified In the same election the Parti conservateur du Québec took 12.91% and won nothing, and the Parti québécois took 14.61% and won three seats while the Parti libéral took 14.37%, a quarter of a point less, and won twenty-one.verified Eighteen seats separated two parties separated by one vote in four hundred.

So the party that in 2022 received seventy-two per cent of the seats for forty-one per cent of the votes received zero per cent of the seats for thirteen per cent of the votes in 2026. Both results came out of the same box. Neither is a malfunction.

§04 · What the ballots could not say

Zoom in one level, from the province to the riding, and the same discarding appears 127 times.

One hundred of the 127 members elected on Monday were elected by less than half the voters in their own riding.verified Sixty-five of them were elected on less than 40%. Twenty-seven were elected on less than 35%. The lowest winning share in the province was 26.03%, in Saint-François.verified In each of those ridings the arrangement is lawful, the count is honest, and the member represents a district in which roughly three voters in four marked someone else.

Province-wide, 2,496,312 votes, or 58.29% of all valid votes cast, went to candidates who lost.verified Add the surplus piled on winners beyond what they needed to win and 3,328,978 votes, 77.73% of everything cast, elected nobody.verified This is not a scandal and it is not unusual; it is the ordinary yield of counting in piles. It is worth stating plainly only because of what is said afterward, which is that the result expresses what Quebec wanted. Three quarters of the marks made on Monday express nothing at all in the output. They were spent, and then they were discarded, and the discarding is invisible in the only number anyone repeats.

§05 · The promise

Here is the keystone, and it is the reason this file belongs in this archive rather than in a textbook. The gearing was on the table. It was promised, legislated, debated, voted on in principle, and then left to die, by the party the gearing had just enriched.

In May 2018, before the election, the leaders of the Coalition Avenir Québec, the Parti québécois, Québec solidaire and the Green Party of Quebec signed a transpartisan agreement committing to replace the voting system with a mixed compensatory proportional model.verified The Coalition then won the 2018 election on 37.42% of the vote and 59.2% of the seats.verified On 25 September 2019 its government tabled Bill 39, an Act to establish a new electoral system, introduced by Sonia LeBel, Minister responsible for Democratic Institutions and Electoral Reform.verified The bill proposed a mixed system in which 80 members would still be elected riding by riding and 45 by regional compensation, with the change ratified by a referendum held alongside the 2022 general election. It was adopted in principle on 8 October 2020 by 93 votes to 28.verified

On 28 April 2021 the government announced there would be no referendum in 2022, citing the pandemic while saying it remained committed to change.reported The bill went no further. By the 2022 campaign the premier had settled on a different explanation: Quebecers did not care enough to change the status quo.reported That election returned the Coalition with 90 of 125 seats on 40.98% of the vote, under the system the agreement had promised to retire.

The party that kept the gear was the party the gear was about to grind.

Four years later the same gear read 13.35% and returned nothing. It is tempting to call that justice. It is not justice; it is indifference, which is the whole point. The machine has no memory and no opinion about who promised what. It converted a governing party into a historic majority in 2022 and into an empty bench in 2026 by applying one unchanged rule to two different maps of where people live. The only decision available about that rule was taken off the table by the beneficiary of the rule, and the public was told the reason was its own lack of interest.

§06 · The majority nobody voted for

One more output of Monday's conversion deserves to be read carefully, because it will be read carelessly everywhere else.

The Parti québécois and Québec solidaire are both sovereigntist parties. Between them they took 40.45% of the vote and 68 of 127 seats, a seat majority of the National Assembly, the first combined sovereigntist majority since 1998.verified The Parti québécois leader campaigned on holding a referendum on independence during a first mandate, and said the commitment held in a minority as in a majority.reported

Both halves of that need keeping apart. The seat majority is real, lawful, and a fact about the legislature. It is not a measurement of support for sovereignty, and it cannot be converted into one. Four voters in ten chose one of those two parties, for reasons that include sovereignty and reasons that have nothing to do with it; the fifty-three per cent is what the converter returned, not what anyone said. On the other denominator, the one that counts every elector entitled to vote, the two parties together drew 26.66%, and the governing party drew 18.47%.verified

This matters beyond arithmetic because a referendum is a different instrument with a different denominator. A seat majority manufactured by piles is a mandate to govern. It is not, and no number on Monday's sheet is, a measurement of how many people in Quebec want to leave Canada. The habit of reading the second off the first is exactly the move this archive files. The companion instrument at Read the Denominator drags the two thresholds apart and shows how far they sit from each other.

§07 · Why a swing reads as accountability

Now the launder itself, which is not in the arithmetic but in what is said over it.

The morning after a result like Monday's, the vocabulary is fixed and it is the same in every language: the voters have spoken, a clear message was sent, they threw them out. Each phrase treats the size of the swing as the strength of the public's feeling. The swing is not a measure of feeling. A shift of a few points in where votes sit can move dozens of seats, and a shift of dozens of points can move none. In 2022, 12.91% of Quebeckers voted Conservative and moved nothing at all. In 2026, 20.67% did, and moved nineteen seats. The feeling tripled and the output multiplied by infinity, because the first number was zero.

What alternation launders is the absence of a decision about the machine. A system that produced the same government forever would be questioned. A system that visibly ejects governments looks self-correcting, and the ejection is offered as proof that it works. So the wheel turns, and each turn consumes the question. Frustration is real, the new vehicle is real, the landslide is real, and in four or eight years the frustration comes back and a different party is loaded into the same chassis. None of those turns is a decision about the chassis. The only vote ever scheduled on the chassis itself was cancelled in April 2021.

The cycle has a second half that the arithmetic cannot reach but that every voter recognises. A party arrives on a landslide it did not earn in votes, governs with a majority larger than its support, declines over a term, and is ejected by a swing larger than the shift in opinion that caused it. Each stage is amplified. Nothing in the amplification corresponds to anything the public did, and all of it is reported as what the public did.

There is a companion to this file that says the same thing in four minutes and a borrowed melody. A Québécois folk round from the fifteenth century has the wolf, the fox and the hare dancing; change one verb and they vote, and nothing else in the song has to move. Edition · The Same Song prints the record, checks the two counts inside it, and sets the clock against every minority government Quebec has ever had.

This is the same move the archive files at Edition · The Menu one step earlier in the process. There, the field of choices is narrowed before election day and the result is stamped as a mandate. Here, the choices are whatever they are and the arithmetic afterward is stamped the same way. In both cases the stamp certifies something the machinery produced, and the machinery is never the thing on the card.

§08 · What this Edition is not

It is not a claim of fraud. The count was honest, Élections Québec reported every poll, and the members elected hold their seats properly. Any reader arriving here looking for a stolen election is in the wrong archive.

It is not a claim that the Coalition Avenir Québec was wronged. A party that loses 27.6 points of support has been rejected by the public on any reading, and the question of whether 13.35% should buy nothing, nine seats, or seventeen is a question about the rule, not about that party's deserts. The same rule paid the same party seventy-two per cent of a legislature four years earlier.

It is not an argument that proportional representation would have produced a better government, or a safer one, or a happier province. Every voting system converts; each converts differently; each has a case against it, and coalition bargaining relocates decisions to rooms voters do not see. The structural point survives whichever system you prefer: the conversion is doing work that is attributed to the public, and in Quebec the one scheduled opportunity to decide the conversion was withdrawn by the government that benefited from it.

And it is not a claim about sovereignty. A sovereigntist seat majority is a fact about seats. What Quebeckers want on that question was not asked on Monday and is not answered by Monday's sheet.