The aftermath: utterly defeated

The election is over, with all ridings declared. Now is a time to take a look at how the projection ranges matched up to actual results:
PQ     25-30% [actual 25.4] (-1.6) 32-50 seats [actual 30] (-10)
PLQ    37-42% [actual 41.5] (+1.5) 60-78 seats [actual 70] ( = )
CAQ    21-25% [actual 23.1] (+0.6)  8-18 seats [actual 22] (+ 9)
QS      7-10% [actual  7.6] (-0.4)  2- 3 seats [actual  3] (+ 1)
ON      0- 1% [actual  0.7] (-0.1)  0- 0 seats [actual  0] ( = )
PVQ     0- 1% [actual  0.6] (+0.1)  0- 0 seats [actual  0] ( = )
Other   0- 2% [actual  1.1] (+0.2)  0- 1 seats [actual  0] ( = )
Black is for within range, red for out of range, differences calculated from median.

As you can see, the polls (and hence, the vote share projection) did extremely well, with everything falling inside the margins. The Liberal majority was also very accurately predicted, with exactly 70 seats. The third QS seat was also within range.

Clearly, however, errors were made with the PQ and CAQ seat projection. Last week I said, "As for the CAQ, incumbency could really be a huge factor for them, especially now that they're on a clear upswing. It's highly likely that the CAQ will save more of their seats than most people expect." It was with this reasoning that I had the CAQ at 13 seats, or between 8 and 18, even as other predictions had them at a maximum of 11. It turns out even that was not enough.

Not only did they save their seats, they actually increased their seat count over last election, even while losing several seats to the Liberals. The size of the PQ collapse allowed Legault to pick up more than enough seats to make up the difference. This is because, even though the CAQ actually dropped 4 points of the popular vote from 2012, the PQ lost nearly 7 points, even worse than expected.

Perhaps, however, the greatest indicator of the magnitude of the PQ defeat was Pauline Marois' loss in her own seat by nearly 800 votes. Her resignation was swift, but notably, Pierre Karl Peladeau was first to make a speech broadcast on television and offer congratulations to Couillard. Words of congratulations to another party leader are normally only spoken by a party leader. Was Couillard signalling his intentions to replace Marois?

There are painful days of soul-searching ahead for the PQ, with a potentially divisive leadership campaign between the traditional unionist wing and the possible "PKP wing" of the party. In this light, I would not be surprised if Legault and the CAQ takes the de facto role of Official Opposition, both in legislature and the eyes of Quebecers, in the following months.

Sovereignty is down, and Canada can breathe a sign of relief. Now the question is, is it out?

Quebec final call & scenarios

Forum has, once again, come out with a shocking poll at the last minute. It shows the PQ vote collapsing (24) and the PLQ gaining (44). The CAQ is again much higher (23) and is nipping at the heals of the PQ. Angus Reid also did a poll, their first in Quebec in several years now, showing a similar close race between the CAQ (25) and PQ (27), but with a much lower Liberal score (39).

With these figures, the PQ once again loses seats in the projection:
PQ     25-30% [median 27.0] (1.4) 32-50 seats [median 40] (5)
PLQ    37-42% [median 40.2] (0.9) 60-78 seats [median 70] (1)
CAQ    21-25% [median 22.5] (1.6)  8-18 seats [median 13] (4)
QS      7-10% [median  8.0] (1.3)  2- 3 seats [median  2] (=)
ON      0- 1% [median  0.8] (0.2)  0- 0 seats [median  0] (=)
PVQ     0- 1% [median  0.5] (0.1)  0- 0 seats [median  0] (=)
Other   0- 2% [median  0.9] (0.1)  0- 1 seats [median  0] (=)

There are several alternate scenarios which could play out, by decreasing likelihood. The main projection is, of course, still the most likely scenario.

1) Angus Reid was the top performer in the many federal and provincial elections. They're right again, and the CAQ vote ends up very close to the PQ:
PLQ 65-71, PQ 33-39, CAQ 16-20, QS 2-3
The PQ edge in efficiency keeps them well ahead in seats.

2) Liberal ballot box bonus, a la 2012 (or, Forum is right):
PLQ 82-88, PQ 24-30, CAQ 9-13, QS 2
Massive Liberal majority, PQ disaster. Liberal strength keeps CAQ down despite vote over 20%.

3) The PQ completely collapses, CAQ carries momentum to finish ahead (in votes). In this situation, the Liberals also drop a bit:
PLQ 59-65, PQ 29-35, CAQ 25-31, QS 2-4
Liberal majority in question, as well as who gets the Official Opposition. PQ weakness also opens up a third and fourth seat to QS. Most exciting situation.

4) Completely defying momentum and trends, the PQ recovers (obviously least likely):
PLQ 53-59, PQ 51-57, CAQ 11-15, QS 2-3
Government formation impossible to predict, CAQ important with balance of power.

Because of the whirlwind end to the campaign, and the CAQ rebounding to dark horse status, there could be a lot of surprises tomorrow night. Once again, we are reminded that campaigns DO matter.

The Politically Uncorrect final call: LIBERAL MAJORITY of 15.

The closing stretch: is the Liberal advantage as solid as it seems?

The campaign is drawing to a close, and all the new polls are pointing at the same thing: a large PLQ lead with the CAQ on the rebound. This projection will likely be the second last one, with the final one coming on Monday morning or so, with all the final adjustments made.

The numbers are here:
PQ     27-31% [median 28.4] (6.2) 41-53 seats [median 45] (8)
PLQ    37-41% [median 39.3] (2.1) 61-75 seats [median 69] (3)
CAQ    19-23% [median 20.9] (7.4)  6-15 seats [median  9] (5)
QS      7-11% [median  9.3] (1.0)  2- 3 seats [median  2] (=)
ON      0- 1% [median  0.6] (0.2)  0- 0 seats [median  0] (=)
PVQ     0- 1% [median  0.6] (0.1)  0- 0 seats [median  0] (=)
Other   0- 2% [median  0.8] (0.2)  0- 1 seats [median  0] (=)

Now is the time to investigate whether the Liberal lead really is as strong as it looks. On paper, a 10+ percentage point lead seems somewhat insurmountable. Alison Redford overcame a small 2pp lead in the polling average, and even Christy Clark "only" overcame an 8pp margin.

The first thing to look at is "likely voters." As many elections have shown us, the people who answer polls may not be the same people who vote. That said, both Ipsos Reid and EKOS have included a "likely voter" formula, and neither shows a bounce for the PQ. I've used their "likely voter" numbers in the poll projection either, so hopefully turnout won't throw this off as far as previous elections.

Looking at recent trends, the PLQ is clearly stagnant or even declining slightly from their peak position. Two weeks ago, they were forecast between 39-44%. Now, that has decreased to 37-41%. A Couillard majority is NOT guaranteed.

However, in the same period, the PQ has had a disastrous stretch, from a forecast of 32-37% to just 27-31%. Those ranges don't even overlap! The only thing that the PQ can hope for now is an incumbency bonus. Usually, incumbent members or governments get a bounce at the ballot box, and they perform better than polls suggest. I may tweak the numbers slightly to account for that in the final update. At this point, though, it looks like the Liberal lead is fairly safe. It would take a small miracle for Marois to pull out a victory now.

As for the CAQ, incumbency could really be a huge factor for them, especially now that they're on a clear upswing. It's highly likely that the CAQ will save more of their seats than most people expect. The same thing happened with their predecessors, the ADQ, in the 2008 election. The projection accounts for that, and so the CAQ seat count is a bit on the high side. That number might rise further by Monday. Legault's second debate performance really saved his hide.

For now, though, Politically Uncorrect projects a LIBERAL MAJORITY of 13.

Week 5: lose-lose

With attack-style politics thoroughly entrenched in Canada, politics is a blood sport at best, and an all-out war of attrition at worst. Elections are seldom won, but often lost. Such is the logic behind the saying, "oppositions can't win elections, governments lose them."

And yet, oppositions can most certainly lose elections as well. Most recently, Adrian Dix lost the BC election by waffling over Kinder Morgan, opening to door to vicious attacks by the Liberals on his economic platform. Even "winners" can also lose. Marois herself had to settle for a razor-thin 4 seat minority last time because of a lacklustre campaign, even while Charest was well and truly f**ked by the Charbonneau commission on corruption.

In fact, the last time someone actually won an election was when Mulroney fought the 1988 election over NAFTA. His record-setting 211 seats in the 1984 election? That was all Turner making gaffe after gaffe.

So, it's not about winning, at all. Rather, it's all about not losing.

And that brings us to Quebec campaign. The question this week: who can survive?

Couillard's performance in the second debate can be described by one word: limp. He was attacked on all sides on the question of integrity. Perhaps the person who swung hardest was Legault, who has been bleeding support to Couillard over concerns of a PQ majority.

Meanwhile, a bombshell dropped today when allegations surfaced that Marois' husband, Claude Blanchet, accepted illegal donations for her leadership bid from a donor who wanted access to Marois. Although she's strongly denied this, the shadow has already been cast over her integrity. What's worse is, she's been trying all campaign to pin the integrity issue onto Couillard. This could make her look like a crook and a hypocrite.

There has been a lack of polling last week, save for a Leger poll confirming the situation before these two events. Expect the next batch of polls to show declines for both Marois and Couillard. Entering the final week of this campaign, whoever drops less is key.